Self-scan stress in Sainsbury’s. And why don’t UK supermarkets use electronic shelf labels?

Posted on By Mark Wilson

Almost every Thursday morning, before I start work, I visit the town market to buy food. After that, I do the weekly supermarket shop. Most people can understand me shopping locally and supporting the market. The question I’m sometimes asked is why I don’t do the supermarket shop online? It’s partly because I’ve learned that the store is well-stocked on a Thursday morning and I can do the weekly shop in 20-30 minutes. There’s also an element of dissatisfaction with previous online supermarket shopping experiences.

I mostly shop at Sainsbury’s. There are some items that we get from Lidl in the next town (though there’s an Aldi locally now, so that may change) and I have to go to Tesco or Waitrose for some other items because the local Sainsbury’s is too small. I also use Costco. Basically, I know what I can get where, at what price/quality.

“SmartShop”

In Sainsbury’s, I use the SmartShop self-scanning technology. According to the Sainsbury’s website:

“SmartShop is the new way to shop at Sainsbury’s. Just scan, bag and go, it couldn’t be easier!”

I started to use this a few years ago, when Sainsbury’s ran a campaign to encourage its use. Then, just a few weeks ago, some tills were removed in our local store to enlarge the self-checkout area. I’ve also switched from using the app on my phone to an in-store handset as I found the barcode scanning to be more efficient.

Random checks

A few months ago, almost every shop was being selected for a “random” check. Sainsbury’s explains that:

“Sometimes customers can double scan an item by mistake, or an item might end up in your trolley that hasn’t been scanned properly. So from time to time we might ask you to have your shopping re-scanned by one of our colleagues.

These rescans are random and they’ll only happen at checkout.”

These were annoying (as it was a regular occurrence), but understandable, until one time the entire shop had to be re-scanned. One of the advantages of self-scanning is that you can carefully place your items in bags so they are not damaged. I watched as my items were re-scanned and roughly repacked for me. I took a deep breath and walked away.

I understand why stores do this. Shoplifting is a huge issue nationally, thought more of a problem in some stores than others. But this policy on self-scans is effectively saying “we think you might be stealing from us”. There’s no apology when no theft is found.

There is an argument that self-scan is also a cost saving measure for supermarkets. That needs to be weighed up against the shrinkage and the customer experience. Some stores simply won’t install self-scan in certain areas, because of the risk.

The “random” checks stopped for a while but today, I was selected again. It’s fair to say that I did not respond well. In fact, I was enraged. 12 September is not a great day for me (it would have been my late Father’s birthday) but I honestly don’t know if that was a factor in my anger when a full re-scan was required. I insisted on speaking to a manager – I don’t believe the scans are random and I’m sure there’s some pattern recognition on my shopping habits via my Nectar card. The last couple of weeks’ shopping was small (with one teenager away from home). This week I spent more, including alcohol, and it felt like I was being singled out.

Pricing errors

The irony is that, after the store re-scanned all my shopping, I found mistakes in their pricing! Far from me adding to Sainsbury’s shrinkage bill, they were not passing on advertised savings to customers.

Readers will probably be familiar with the concept of discounts for loyalty card holders. Tesco has Clubcard, Sainsbury’s has Nectar, other retailers have their own schemes too. These are controversial for some, but I’m comfortable accepting that I trade my data for cash. After all, I give data about my habits away all the time on the Internet, using “free” services (if you’re not paying for the product, you are the product).

I found that Sainsbury’s had not passed on a Nectar discount on one of my items. Furthermore, because the ePOS system was not configured with the correct price, it would presumably have been overcharging every customer who bought that item and used their Nectar card.

Then, later in the day, I spotted that some of the personalised Nectar offers from a SmartShop scan were not passed to me when I’d had the full shop scanned through a normal till. Those offers were actually a reason for me to buy multiple items, rather than just one. They had increased the volume of the sale, but I’d ended up paying the full price.

Both of these mistakes were corrected by staff but they shouldn’t have happened.

In summary, when Sainsbury’s systems suspected I might have been shoplifting, it actually turned out that they owed me money.

Teenage kicks

I started my working life in retail. As a teenager, I worked for Bejam (now Iceland), and then a few years at Safeway. It was mostly stacking shelves but also warehouse work and checkouts when the store was busy. I saw the change from manual pricing to ePOS with barcodes, and I worked on a number of store openings and refits. After I decided to go to Polytechnic instead of joining Safeway’s management programme, I came back in my holidays and worked night shifts. That period of my life taught me quite a lot about supermarket retail and, fundamentally, not much has changed since. Of course, there have been some developments – like just-in-time deliveries replacing in-store warehouse space and the creation of digital services such as online shopping and self-scan.

One thing that does seem to have changed though is the checks on price labels. At Safeway in the late ’80s and early ’90s, it was a full-time job to check every price in store and manage changes/promotions. If the shelf edge labels didn’t match the computer then the customers were charged the wrong price. That was taken seriously back then.

This attention to detail seems to be gone. I imagine it was a cost-cutting efficiency (as is self-service). Nowadays, I regularly spot pricing errors in Sainsbury’s and it usually leads to store staff removing errant shelf edge tags. And Sainsbury’s are not alone – the local Co-op and OneStop stores seem to have similar issues.

Electronic shelf labelling

So, why don’t UK supermarkets use electronic shelf labels (ESLs), like those seen in continental Europe? I did some basic analysis and it seems that early trials were inconclusive, with concerns around cost, technology and operational challenges. So, just like any IT system really.

On the other hand, the benefits include efficiency, dynamic pricing, customer information and sustainability. The Grocer reported in 2021 that ESLs were making a comeback but I’ve not seen much evidence to suggest it’s happening quickly.

So what might ESLs cost for a store like the one where I shop, which was only built 5 years ago?

My local Sainsbury’s store cost £3.3m to build and is 1610 square metres in size. A few prompts to an AI assistant has told me that:

  • A store this size can be expected to stock 20-25,000 product lines.
  • The cost of ESLs can vary depending on the brand and features but an investment for 10,000 lines would be around £50-80,000.

So, about £125-200,000 for a store this size (between 3.7 and 6% of the £3.3m budget) to have accurate pricing in store.

No business case?

The thing is, that, in addition to my teenage shelf-stacking, I have some IT experience of working in retail. When I was at Polo Ralph Lauren in the early 2000s it was a lot easier to justify application spend than infrastructure. If IT spend doesn’t add to the bottom line, then the business case is unlikely to be approved. And if stores make more money from advertising offers that are not applied, why would they invest in a system to display accurate pricing?

Call me a cynic, but could that be the real reason why UK stores haven’t invested in electronic ticketing?

Featured image: author’s own.

The Enterprise Architecture Stack

Posted on By Mark Wilson

Over the years, I’ve written several posts about IT architecture. Whilst it seems that there is an increasing trend to call experienced IT folks “architects”, one of my core beliefs is that Enterprise Architecture is not the same as “architecting” IT at enterprise scale. Yes, creating an IT architecture that will scale to support a global organisation with thousands of users is “enterprise scale” – but it’s not Enterprise Architecture.

So what is Enterprise Architecture?

Like so many things in life, an illustration can really help describe a point. And, a few years ago, I came across an excellent Enterprise Architecture diagram from Dave Clark and Sophie Marshall. You can see it as the featured image at the top of this post and one of the reasons I like it so much is that it’s clear that the technology is only one of several factors in a whole stack of considerations.

I adapted it (under Creative Commons) but the basic premise of the diagram remained the same – step back from the problem and understand the organisation to consider its needs and requirements. We need to know what is needed before we can can consider solutions! Then, we should ask what good looks like. Don’t just dive in with technology.

Let’s take each layer in turn… and you’ll see that, right away, I added another layer at the top.

Purpose

The purpose is about why an organisation exists. It should be straightforward to answer but is hopefully more than “to deliver value to our shareholders”. A Council may exist to provide services (statutory and otherwise) to citizens. A retailer may exist to (make money and) provide the best selection of fashionable clothing at affordable prices. It’s entirely logical that the organisation’s culture will be strongly linked to its business motivations.

Many organisations will give an indication of their purpose on their website, or in their company report. For example, the IKEA vision, values and business idea sets out the organisation’s purpose in the form of:

  • A vision: “To create a better everyday life for the many people”; and
  • A business idea: “to offer a wide range of well-designed, functional home furnishing products at prices so low that as many people as possible will be able to afford them.”

Strategy

Strategy supports purpose by providing business ambition and goals – a direction in which to head. Storytelling and visualisation are techniques that can be used to communicate the strategy so that it’s well understood by everyone in the organisation. They can also help others who need to work with them (for example business partners). A useful tool for defining business strategy is the Business Model Canvas, based on the book by Alexander Osterwalder and Yves Pigneur.

Looking briefly at visualisation, Scott Berinato (@ScottBerinato)’s 2016 article for Harvard Business Review on Visualizations [sic] That Really Work stresses the need to understand the message you want to convey before you get down into the weeds. This blog post is a case in point – I want to show that Enterprise Architecture is much more than just technology. And I found a good visualisation to illustrate my point.

As for storytelling, I’ve seen some fascinating presentations over the years on how to tell a good story to bring a presentation to life. One of the most memorable was at a Microsoft MVP Event in 2017. Tony Wells used this example of how we tell stories to children – and how we (too) often communicate at work:

(I’m still practicing my storytelling technique, but Hubspot also has what it calls The Ultimate Guide to Storytelling.)

What, Who and How

What we do is a description of the products and services that the organisation offers – the business’ capabilities. These may be the value propositions in the Business Model Canvas but I would suggest they are a little more detailed. Strength/Weakness/Opportunity/Threat (SWOT) analysis can be a useful tool here too for identifying what could be done, though the emphasis is probably more on what is currently done, for now.

Who we do it for is about the consumers of the organisation’s products and services – understanding who the “users” are. Tools might include stakeholder maps and matrices, empathy maps, personas.

How it’s done is about understanding the methods and processes that deliver “the what” to “the who”. Journey maps, process flow diagrams, storyboards and SWOT analysis can all help.

Who does it is about the people, where they are located, and how the organisation is structured. In a world of remote and hybrid working it’s even more relevant to understand the (human) network and how it works.

Software, data and technology

Only after we’ve understood “the Business layers” (purpose, strategy and the what, who and how) can we move onto the IT. And that IT is more than just infrastructure:

  • The data models that support this. (There may a discussion to be had there about data, information, knowledge and wisdom but that’s a topic in itself.)
  • The software applications that are used to access that data.
  • And the underlying technology infrastructure.

Why is this important?

For many years, I was part of and then managed a team of people who were labelled “Enterprise Architects”. During that time, I argued that the term was aspirational and that most of the work we did was Solution Architecture. Maybe that was splitting hairs but we rarely got the chance to drive strategy, or to get involved in designing the organisational structure. Whilst we were experienced at IT, we still operated at the lower levels in the stack: business requirements driving software, data and technology decisions. We wanted to become trusted advisors, but for the most part, the work we performed for our clients was transactional.

My colleague Ben Curtis (/in/BenCurti5) has an excellent analogy built around perception and perspective. I hope he won’t mind me borrowing it:

  • Perception is about what meets the eye. Imagine you’re walking through a forest and come across a single tree. Your first impression of that tree – its size, shape, colour, and surroundings – is your perception.
  • Perspective is seeing the Forest and the Trees. Now, let’s say you climb to the top of a hill and look down at the entire forest. Suddenly, you see how all the trees are connected, how the sunlight filters through the leaves, and how animals move through the undergrowth. This bigger view – the perspective – gives you a deeper understanding of the forest as a whole.

Whilst this can be used to show the difference between an individual system and the complete view of an IT environment, I’d suggest that its also about how the IT environment is part of something much larger – an organisation of people and processes, supported by technology, that exists with a purpose and a strategy to make it happen. And that, is the Enterprise Architecture.

Related posts

Here are some posts I’ve written previously on IT architecture. I think this is the first time I’ve properly outlined what Enterprise Architecture means though:

Featured image: The Enterprise Architecture Stack, by Dave Clark and Sophie Marshall [source: Dave Clark on LinkedIn]

Watercooler moments, and hybrid work

Posted on By Mark Wilson

In last week’s weeknote, I asked a question:

“Much is made of “watercooler moments” as a reason to return to the office (RTO). Well, is there any reason that such moments can’t happen outside the office too?”

I wrote about “coffees” (as a metaphor for meeting up with no agenda), but only this afternoon I witnessed a “watercooler moment” away from the office. Two of my colleagues were on a Teams call, and they discovered that they live only a few streets away from one another. Actually, as the call went on, I realised that 80% of the attendees were in/around the same city (I was the exception).

But what also became apparent to me is how these four people in the same city had different needs – and that travelling to a city centre location was almost easier for me (80 miles away) than it was for some of them!

This is where the subject of this blog post shifts to hybrid work. And it seems that the thing to remember about hybrid work is that what works for one does not necessarily work for another.

It’s complicated

On the one hand, we want to nurture a culture, and to get people working as a team. One way to do that is to co-locate them. But we have distributed teams too. Regionally, nationally or globally. There’s limited value in going to the office if you won’t actually be in the same place as your colleagues. Conversely, there’s an view that you might meet people from a broader cross-section of the company. That holds some merit.

I planned to be in an office tomorrow, then my diary filled up and it looks like I will spend half the day on the same Teams calls I would have attended from home. Is that a good reason to be in the office?

There’s also the view of productivity. Being present in an office isn’t the same as doing good work. Some find it easy to work at home. Others struggle to get motivated. Some find it easier to work in an office. Some struggle with the noise and disruption. Some people have better IT at home (e.g. a faster connection to the Internet).

Some people will say “you used to come to the office before Covid”. But did we? Working patterns have been changing for a while. Even before the pandemic, I was at home or on client sites most of the time. I was rarely at a company location.

And there seems to be an assumption by some that employees took remote or hybrid jobs as if it was some kind of favour, and can now absorb the time and costs associated with commuting. I’ve been contractually based from home since 2005. If I travel to an office, it costs the company money, not me. That may work for a professional services business charging me out on a fee-earning basis, plus expenses. But less so when it’s “everyone into the office x days a week”.

I do feel for those who are starting out or early in their careers. It must be tricky when you don’t see the people you work with. I learned by observing others, and by being shown what to do. We need to find new ways to do that and to nurture people’s growth.

But, for those with caring responsibilities, the flexibility of working from home means they can continue delivering value to the organisation whilst fulfilling the needs of those they care for. As a society, we’re living longer. Those of us in our middle age are sandwiched between the needs of our children and those of our aging parents. And that’s not considering the changes in our ability to work that come with this time of life, for example adapting to deal with the menopause (that’s a whole blog post in itself).

We don’t want to lose good skills to when we really just need to accommodate different people’s needs.

And then there’s the green angle. “How can we be more environmentally-sensitive?”, asks the person responsible for the organisation’s corporate social responsibility (CSR) goals. Perhaps, by travelling less?

It’s about autonomy

I’ll be the first to admit that 100% remote work doesn’t work for me. And I get a lot from being with others to collaborate on something. Earlier this month I did travel to meet a colleague and “write on the walls”. We got a lot of value from that meeting.

And that’s the nub of it. Not blanket RTO edicts, as I’ve heard mentioned in multiple organisations recently, but autonomy and flexibility. Or, as one Gartner Analyst put it:

“The key to hybrid work is allowing employees the autonomy to work when they want and where they want.”

Gavin Tay, Gartner Inc.

Provide a place for people to meet and collaborate. Provide a place with a desk and decent Internet connection for those who don’t have a suitable workspace at home. Encourage people to come together. But be clear what it is you are trying to achieve. If the answer involves filling up rows of hot desks, then you may want to think again.

Related posts

It seems this is a topic I keep returning to (I clearly have Opinions on the topic), so here’s some posts I wrote earlier:

Featured image by Israel Andrade on Unsplash.

Learning to be intelligent about artificial intelligence

Posted on By Mark Wilson

This week promises to be a huge one in the world of Artificial Intelligence (AI). I should caveat that in that almost every week includes a barrage of news about AI. And, depending which articles you read, AI is either going to:

  • Take away all our jobs or create exciting new jobs.
  • Solve global issues like climate change or hasten climate change through massive data centre power and water requirements.
  • Lead to the demise of society as we know it or create a new utopia.

A week of high profile AI events

So, why is this week so special?

  1. First of all, the G7 nations have agreed a set of Guiding Principles and a Code of Conduct on AI. This has been lauded by the European Commission as complementing the legally binding rules that the EU co-legislators are currently finalising under the EU AI Act.
  2. Then, starting on Wednesday, the UK is hosting an AI Safety Summit at “the home of computing”, Bletchley Park. And this summit is already controversial with some questioning the diversity of the attendees, including Dr Sue Black, who famously championed saving Bletchley Park from redevelopment.
  3. The same day, Microsoft’s AI Copilots will become generally available to Enterprise users, and there’s a huge buzz around how the $30/user/month Copilot plays against other offers like Bing Chat Enterprise ($5/user/month), or even using public AI models.

All just another week in AI news. Or not, depending on how you view these things!

Is AI the big deal that it seems to be?

It’s only natural to ask questions about the potential that AI offers (specifically generative AI – gAI). It’s a topic that I covered in a recent technology advice note that I wrote.

In summary, I said that:

“gAI tools should be considered as assistive technologies that can help with researching, summarising and basic drafting but they are not a replacement for human expertise.

We need to train people on the limitations of gAI. We should learn lessons from social media, where nuanced narratives get reduced to polarised soundbites. Newspaper headlines do the same, but social media industrialised things. AI has the potential to be transformative. But we need to make sure that’s done in the right way.

Getting good results out of LLMs will be a skill – a new area of subject matter expertise (known as “prompt engineering”). Similarly, questioning the outputs of GANs to recognise fake imagery will require new awareness and critical thinking.”

Node 4 Technology Advice Note on Artificial Intelligence, September 2023.

Even as I’m writing this post, I can see a BBC headline that asks “Can Rishi Sunak’s big summit save us from AI nightmare?”. My response? Betteridge’s law probably applies here.

Could AI have saved a failed business?

Last weekend, The Sunday Times ran an article about the failed Babylon Healthcare organisation, titled “The app that promised an NHS ‘revolution’ then went down in flames”. The article is behind a paywall, but I’ve seen some extracts.

Two things appear to have caused Babylon’s downfall (at least in part). Not only did Babylon attract young and generally healthy patients to its telehealth services, but it also offered frictionless access.

So, it caused problems for traditional service providers, leaving them with an older, more frequently ill, and therefore more expensive sector of the population. And it caused problems for itself: who would have thought that if you offer people unlimited healthcare, they will use it?!

(In some cases, creating friction in provision of a service is a deliberate policy. I’m sure this is why my local GP doesn’t allow me to book appointments online. By making me queue up in person for one of a limited number of same-day appointments, or face a lengthy wait in a telephone queue, I’m less likely to make an appointment unless I really need it.)

The article talks about the pressures on Babylon to increase its use of artificial intelligence. It also seems to come to the conclusion that, had today’s generative AI tools been around when Babylon was launched, it would have been more successful. That’s a big jump, written by a consumer journalist, who seems to be asserting that generative AI is better at predicting health outcomes than expert system decision trees.

We need to be intelligent about how we use Artificial Intelligence

Let me be clear, generative AI makes stuff up. Literally. gAIs like ChatGPT work by predicting and generating the next word based on previous words – basically, on probability. And sometimes they get it wrong.

Last week, I asked ChatGPT to summarise some meeting notes. The summary it produced included a typo – a made-up word:

“A meeting took tanke place between Node4 & the Infrastructure team at <client name redacted> to discuss future technology integration, project workloads, cost control measures, and hybrid cloud strategy.”

Or, as one of my friends found when he asked ChatGPT to confirm a simple percentage calculation, it initially said one thing and then “changed its mind”!

Don’t get me wrong – these tools can be fantastic for creating drafts, but they do need to be checked. Many people seem to think that an AI generates a response from a database of facts and therefore must be correct.

In conclusion

As we traverse the future landscape painted by artificial intelligence, it’s vital that we arm ourselves with a sound understanding of its potential and limitations. AI has often been regarded as a silver bullet for many of our modern challenges, a shortcut to progress and optimised efficiency. But as we’ve explored in this blog post – whether it’s the G7 nations’ principles, Microsoft’s AI Copilot, or a fallen Babylon Healthcare – AI is not a one-size-fits-all solution. It’s a tool, often brilliant but fallible, offering us both unprecedented opportunities and new forms of challenges.

The promises brought by AI are enormous. This week’s events underscore the urgency to familiarise ourselves with AI, acknowledge its potential, and intelligently navigate its limitations. From a set of AI guiding principles on a global scale, to raising awareness on gAI, and analysing the role of AI in business successes and failures – it’s clear that being informed about AI is no longer an option but a necessity.

gAI tools, while they are transformative, need to be used as assistive technologies and not as replacements for human intellect and expertise. Embracing AI should not mean renouncing critical thinking and caution. So, as we interact with AI, let’s do it intelligently, asking the right questions and understanding its strengths and limitations. We need to be smart about using AI, recognizing both its potential and its constraints. This will enable us to harness its power effectively, while avoiding over-reliance or the creation of new, unforeseen problems.

It’s time we stop viewing AI through a lens of absolute salvation or doom, and start understanding it as a dynamic field that requires thoughtful and knowledgeable engagement. Evolution in human tech culture will not be judged by the power of our tools, but by our ability to skillfully and ethically wield them. So, let’s learn to be intelligent about how we use artificial intelligence.

Postscript

That conclusion was written by an AI, and edited by a human.

Featured image: screenshot from the BBC website, under fair use for copyright purposes.

Breaking down and planning big tasks (e.g. for exam revision)

Posted on By Mark Wilson
This content is 1 year old. I don't routinely update old blog posts as they are only intended to represent a view at a particular point in time. Please be warned that the information here may be out of date.

In common with many young people in households across England and Wales, for the last few weeks, both my sons have been taking their end-of-school exams (Scottish schoolchildren finished theirs a few weeks ago).

My youngest son had more than 20 exams for his GCSEs; my eldest had eight for his A-Levels. On the lead up to these (and between them), there was a lot of revision to be done.

Creating a plan

Back when they were sitting mock/trial exams, we noticed that schools don’t teach young people how to plan. At least not based on my experience of two state secondary schools in Milton Keynes and Northampton. They might provide a list of topics, or even a per-subject “revision timetable” but my wife realised fairly early on that our sons could just see some dates, and a massive task ahead of them.

So, we sat them down, and helped to worth things through. Using Excel of course (other Project Management tools are available, but probably overkill)!

  1. First, look how many weeks there are until the exams. The days are your columns. Use borders/shading to see the weeks and weekends.
  2. Then, look at the subjects you need to cover. Those “swimlanes” are the rows. Break each swimlane into 3 rows: daytime; after school; evening.
  3. Then block out the time for actual school, part-time work, sports activities, holidays, etc.
  4. You can now see the amount of time that’s available for revision/study and populate each spare block with one or more topics to cover within each subject row.

I expected some push-back, but was amazed how positively they took on the advice (to the extent that they seemed to work it out for themselves and created their own plans when it came to the final exams).

It’s simple project management!

What we taught them to do was effectively basic project management. It’s effectively using a Gantt Chart to illustrate the schedule for completing a bunch of tasks, along with the resource availability and constraints.

This is a life skill but also a business skill. It amazes me that this isn’t taught in schools (even pretty good ones). Or perhaps it is, but it’s lost in the teenage air of nonchalance!

Featured image by Eric Rothermel from Unsplash.

This is what happened when I asked ChatGPT to write a blog post…

Posted on By Mark Wilson
This content is 1 year old. I don't routinely update old blog posts as they are only intended to represent a view at a particular point in time. Please be warned that the information here may be out of date.

In recent weeks, I’ve been watching the buzz of activity about large language models and generative AI. So I decided to see what would happen if I asked Open AI’s ChatGPT to write a blog post for me. I gave it the following prompt:

Write a blog post of up to 800 words that discusses the use of generative AI tools like ChatGPT to automate human activities Use writing a blog post as an example. Include some commentary about the potential social impacts of using AI to generate content. Also discuss whether there are any concerns that need to be considered - such as whether AI-generated content is accurate.

“Generative AI tools like ChatGPT have been making waves in the field of automation and artificial intelligence. These tools have the ability to generate human-like text, helping to automate certain human activities such as writing, customer service, and more. While the potential benefits of these tools are clear, there are also concerns about their impact on society and the accuracy of the content they produce.

One of the most significant benefits of generative AI tools like ChatGPT is their ability to automate repetitive or time-consuming tasks. For example, a blogger who regularly posts new content may use ChatGPT to generate article outlines, saving time and effort. Customer service teams can use generative AI tools to provide quick and accurate responses to frequently asked questions, improving the overall customer experience.

Another benefit of generative AI tools is their ability to create content in multiple languages, allowing businesses to expand their reach into new markets. These tools can also be used to generate personalized content for individuals, such as email newsletters or social media posts, which can improve engagement and lead to higher conversions.

However, the use of generative AI tools also raises concerns about the potential social impacts of using AI to generate content. One concern is that these tools could lead to a loss of jobs, particularly in industries where writing and content creation are essential. While automation has always been a part of technological progress, there is a risk that AI could replace more jobs than it creates.

Another concern is that AI-generated content could be used to spread misinformation and propaganda. With the ability to create convincing, human-like text, there is a risk that malicious actors could use these tools to create fake news stories or manipulate public opinion. This could have serious consequences for democracy and social stability.

Accuracy is another concern when it comes to generative AI tools. While these tools are designed to create human-like text, they may not always produce accurate or reliable information. This could be particularly problematic in fields like journalism or science, where accuracy is essential. As AI tools become more sophisticated, it will be important to ensure that the content they produce is factually correct and trustworthy.

Despite these concerns, generative AI tools like ChatGPT have the potential to revolutionize the way we create and consume content. As with any new technology, it is important to approach these tools with caution and to carefully consider their potential impact on society. However, if used responsibly, these tools could improve efficiency, accuracy, and creativity in a variety of industries.

In the case of writing blog posts, for example, using ChatGPT to generate outlines or ideas could be a useful starting point, but it is important to remember that these tools are not a substitute for human creativity and expertise. While AI-generated content may be faster and more efficient, it may lack the nuance and personality that comes from a human writer.

To ensure that AI-generated content is accurate and reliable, it is important to use reliable sources of information and to fact-check any content that is produced. This is particularly important when it comes to sensitive or controversial topics, where misinformation can have serious consequences.

Overall, the use of generative AI tools like ChatGPT to automate human activities is a promising development that has the potential to improve efficiency and creativity in a variety of fields. However, it is important to approach these tools with caution and to carefully consider their potential impact on society. With responsible use and careful oversight, generative AI tools could become an important part of our technological landscape, helping us to achieve more than we ever thought possible.”

Is it good enough?

As it happens, the resulting text is quite believable, but it’s not actually a very good post. A different AI, the one from Yoast SEO that helps with the readability of my content, highlighted that there aren’t enough subheadings. Also, more than half the sentences are a bit long (more than 20 words). And it’s using too much passive voice (I do that too, so I guess it could be considered authentic!).

Garbage In Garbage Out (GIBO)

I tried asking ChatCPT to improve its writing, based on the points above but I didn’t get far:

Re-write the above with shorter sentences to make it easier to read, and use sub-headings.

And then:

Try again, and this time use try to use active voice instead of passive voice... and make the subheadings a bit snappier!

It made some changes, but really the way I see tools like this working is as a very quick first draft for review and update by a human writer. We just have to be careful not to let those who don’t understand or who don’t value writing say “just get an AI to write it in a few seconds”.

Featured image by Alexandra_Koch from Pixabay.

Recruitment

Posted on By Mark Wilson
This content is 3 years old. I don't routinely update old blog posts as they are only intended to represent a view at a particular point in time. Please be warned that the information here may be out of date.

The last quarter of 2021 was manic. I’d already “lost” one of my team, who was working his notice period before joining a competitor, when another team member told me he’d decided to follow opportunities outside the company. All of a sudden, my delayed decision to recruit (over concerns about keeping the team busy) became an urgent need to recruit two experienced Solution/Enterprise Architects, just as our workload hit a spike.

With a lack of internal candidates coming forward (often the good technical Consultants want to stay close to tech), I discussed the issue with my senior management team and we advertised externally for two Enterprise Architects.

Finding the right candidates

I’ve recruited before, but this was the first time I’d been a Hiring Manager at risual. In my experience, every company has a different approach to recruitment and practices change over time too. I was fortunate to be working with a fantastic HR Advisor, who helped me specify the role (having an up to date Job Description helped too). Over the next few weeks, adverts went out on Indeed, on LinkedIn, and through various recruitment partners. Slowly, but surely, the CVs came rolling in.

Whilst we all put a lot of effort into creating CVs, they really are just the “foot in the door”. The first sift happened before they even got to me. Of those I saw, I rejected some because they didn’t seem to relate to the role as advertised. Maybe those candidates had the experience, but it wasn’t clear from their CV. I was recruiting for two senior roles and some people seemed to take an opportunistic approach. Other CVs were too long, listing everything the candidate had ever done over an extended period. There’s a fine balance between not enough detail and too much. Just remember that, although you put hours into writing that CV, it may only get a 30 second skim – or perhaps a bit longer if you manage to grab the reviewer’s attention.

I also saw CVs with typos. And I wanted to meet a candidate who looked fantastic but had neglected to include contact details. And, sadly, I saw well-formatted CVs that had been butchered by the recruitment agency’s topping and tailing.

In the end, I decided to meet with around half the candidates whose CVs I’d reviewed. Actually, it was slightly more than that but we had problems contacting at least one candidate (as mentioned above) and others had already accepted roles elsewhere (including one who only told us when we contacted him a couple of hours before his planned interview). This is a fast-moving market and, right now, it’s definitely favouring those looking for a new job over those looking to hire.

With CVs sifted, the remaining process would be an interview with myself and my manager (as the hiring team), some personality and numeric/verbal reasoning tests, and finally an interview with the Chairman and the CEO. The aim was to move quickly – from first interview to offer in around a week.

Interview criteria

I remember the first time I ever took part in an interview (from the hiring side). My then-Manager, a wise Managing Consultant by the name of Mark Locke, told me that it’s quite simple:

  1. Does this person know what they are talking about?
  2. Can I work with them?

This advice still holds true today. It’s pretty obvious, pretty soon, when an interview is going badly. The good ones are pleasurable.

Looking back, I can call out some really enjoyable discussions at interviews that went on to be great hires. One in particular (back in my days at Fujitsu – and before Microsoft Teams) was unavailable when I called his mobile phone (he’d been driving for work and struggling to get a signal at the appointed time) but, after we rescheduled and finally got to speak, my initial impressions were overturned by someone who turned out to be an extremely talented Project Manager with a passion for technology. We went on to work together before his career went from strength to strength. These days he’s a Senior Program Manager at Microsoft and he’ll know who I’m writing about if he still reads this blog.

As a candidate, I’ve written previously about some shocking experiences but my risual interview was different. I was a bit put off when Alun Rogers told me he reads my blog (back in the days when I used to post more!) but it felt like it had gone well as I drove home later. It really did feel like “just a chat” and I enjoyed meeting Al and Rich (Proud). Similarly, when David Smith had interviewed me a few years earlier to join his Office of the CITO at Fujitsu, I just had a feeling that things had gone well.

For my recent hires, I had formal criteria to assess against but my Manager and I had also worked out a set of standard questions to structure our conversations with the candidates. Each hour-long interview had at least that long spent writing up and reviewing the notes, as well as prep time. And all of that had to fit around my management and delivery roles – so November and December 2021 were intense!

What did I learn?

The whole experience taught me a lot. That’s why I’m sitting here, in the gap between Christmas and New Year, writing about it (whilst it’s still reasonably fresh in my mind).

First of all, I realised that the same question can elicit very different responses to the one that was first expected. And, just because someone has a different view to me, doesn’t make them wrong. A gut feel about being able to work with someone is good but, if you only look for people who think like you, it won’t do much for the diversity of the team.

Having said that, for the candidates who looked me up on LinkedIn (and there were some – good interview preparation, I’d say), you don’t have to go far back on my blog to see a post about what I expect to see in an IT Architect…

I also (re-)learned that interviewing is hard. Not only is it demanding from a cognitive perspective but there is a lot of work to do both before and after the interview. And not doing that is not giving the candidate the respect that they deserve – I will always put in the effort.

Interviewing is enjoyable too. No-one wants to see anyone struggle and, as I wrote earlier, I genuinely enjoyed some of the discussions I’ve had in the past and the same can be said for the ones in recent weeks (though I’m not going into specifics here for reasons of confidentiality).

I learned some other things too – things that I can’t write about here because they relate to specific details of individual hires but which were nevertheless valuable in me learning to trust my own judgement (e.g. after having to interview alone, instead of as part of a team).

And I learned that not all the advice given by recruitment partners is correct. Again, I won’t go into details but the right candidates are out there.

I was also intrigued by the personality tests. So much so that I asked if I could do them myself. I completed them before I left for the Christmas break and I’m looking forward to seeing how I compare to the candidates we’ve recruited when I get the reports. Again – I’m not looking for people who think like me… but I am looking how the tests assess me and how that relates to the way I think. It might also be useful to see how middle-aged me compares to younger me.

Looking forward and rebuilding the team

Now, as we go into 2022, I’m really excited to have two new starters joining my existing team, to help shape our future and support the company’s growth plans. As for the guys who moved (or are moving) on – I genuinely do wish them well. I know one is having a great time in his new role and the other has an exciting opportunity lined up. I’d rather we were all working together (new hires and “the old team” together), but I’m also a realist and sometimes the best thing to do is to support people in taking their next steps.

As one former Managing Director used to say when signing off his communications… Onwards!

Featured image by VIN JD from Pixabay.

Step back from the problem and think about what “good” looks like

Posted on By Mark Wilson
This content is 3 years old. I don't routinely update old blog posts as they are only intended to represent a view at a particular point in time. Please be warned that the information here may be out of date.

A few weeks ago, I sat down with a Chief Information Officer (CIO) who has a problem. He’s in the middle of a messy “divorce” (professionally, not personally). He is transitioning services from a shared services agreement with another public sector body to a new managed service. His own organisation’s IT maturity is low. There’s an expectation that the new managed services partner will take on everything (except it’s not in a state that is ready to take on). And the shared service provider is both making transition difficult (preserving its revenue stream) whilst ramping up the price to carry on providing services. The divorce metaphor is very apt. 

I was brought in (alongside a colleague with relevant sector experience) to help smooth the pain. I needed to understand what’s holding up the process – why is it so difficult to provide basic information for the managed services provider to take on the service? What are the gaps? How quickly can they be filled? And what is needed to move to the next stage? 

It’s not my usual role, but I’ve been around this industry long enough to be able to take a step back, look at the problem, and try to work out what “good” looks like. 

The challenges

The CIO presented me with two challenges: 

  1. Visibility – of what’s happening. What will be done by when and how far off the target is the transition?
  2. Passiveness – don’t just sit and wait. Bang down some doors and ask for information. If it’s not forthcoming, then flag it. There is no time for delays. 

Searching for a solution

The next day, I was mulling over the issues and I bumped into a friend (on the market in the town where we both live). We went for a coffee, and I told him about my problem (without compromising any confidentiality). My friend has a military background, followed by IT Service Operations and, more recently, security (he’s a Chief Information Security Officer – CISO) so I shouldn’t have been surprised by the advice he gave me. The way he saw it was that there are a bunch of service transition “packages” but the business as usual (BAU) service isn’t complete. Meanwhile the CIO has no visibility and would like to see where things are and the plan for where they will be.

After our conversation my mind was clear. I needed a way to track progress. I wanted a dashboard to tell me the state of each service component or process. Then, the applications, servers and other infrastructure could fall in beneath – but I needed to know there is a service to transition them into. 

There are many problems with dashboards (though the etymology of the word is about protecting riders on carriages from what might be thrown at them from below… so maybe that’s quite appropriate after all). Red/Amber/Green (RAG) statuses can be problematic too (both for cultural reasons and because of accessibility, although that can be overcome with appropriate design). But I didn’t want perfect – I needed functional. At least for the first iteration.

The chosen approach

The Microsoft-focused Solution Architect in me was thinking Power BI but I lacked the skills, time and access to licenses. I needed something that could be developed quickly and updated easily. My initial PowerPoint deck with, “this is what we said we would do”, “This is where we are today” and lots of red, turning amber then green was quickly pushed aside by a colleague in favour of Excel. In fairness, the world runs on Excel – and that’s not necessarily a bad thing. With the addition of a few formulae, some data validation and some conditionally formatted cells, we soon had a dynamic report. It highlights missing information. It highlights support status. It highlights key dates (and missed dates – because I’m also realistic).

Answering the exam question

The summary sheet should answer the CIO’s visibility issue (once it’s securely shared) and constantly pushing for the detail should strike out any perceived inactivity or a lack of initiative.

It’s not innovative, but it is elegant. And it works. 

So I have the tech in place – now for the difficult bit (the part that involves people) – dragging out the missing information to turn cells from red to amber to green. And the good thing is that, based on a meeting yesterday, it feels like there are a bunch of people in the managed services organisation that can see the value and are invested in the solution (they are even adding sheets for extra information – like tracking risks, issues and dependencies). That’s half the battle. “All” I need now is to get the various projects that hold the information on the various applications, servers, etc. to join in.

I may return to this subject with an updated post when everything goes live. Or I may not, for commercial reasons, but here goes… wish me luck! 

Featured image: author’s own.

Not-so-helpful social media “help”

Posted on By Mark Wilson
This content is 3 years old. I don't routinely update old blog posts as they are only intended to represent a view at a particular point in time. Please be warned that the information here may be out of date.

Social media is big business. And almost every major business to consumer (B2C) organisation has at least one account on each of the major social media platforms (at the time of writing, that’s Twitter, Facebook and Instagram but I’m sure it will change over time). 

Unfortunately, there’s a concerning trend starting to emerge – one where the “conversation” is moved to control the brand image. Many brands have set up <brandname>Help accounts for their customer service so that the main brand account is “clean” – pure marketing, untarnished by customers expressing concern about the products and services. Meanwhile, the “Help” account may be operated by a communications agency, simply offering a face and redirecting customers to other channels. 

And that’s where the problem lies. If you want to offer omnichannel support, then you need to meet your customers where they contact you. It’s no good offering “help” on Twitter when all you’re really doing is advising customers to phone your contact centre. That does not help. That’s obfuscation. It’s a blatant attempt to preserve the online image of the brand, whilst offering shoddy customer service. 

So, here’s my plea to brand managers across the UK. If you offer a <brandname>Help account, then make sure it provides real assistance and is not just signposting to another channel. 

I’ll provide an example here, from @KwikFitCS (who responded to my tweet for the main KwikFit account… more on that in a moment), but they are not alone…

Then there’s the issue of the information that <brandname>Help accounts ask for to verify you before they will provide help…

In the example above, @BootsHelp replied to a tweet sent to @BootsUK. And the issue I was reporting was a website problem that was not specific to a single account – the web team could investigate without my personal details. Maybe I should be the one looking for the verification here… not them? That may sound a bit extreme but what’s to stop anyone from setting up a spoof <brandname>Help account and harvesting information from disgruntled customers? (In fairness, the @BootsHelp account has been verified by Twitter, but the @KwikFitCS example earlier was not).

And Boots is not alone – here’s another example from @Morrisons, the UK supermarket chain:

The request went on to a second tweet:

So, come on B2C Twitter. You can do better than this. How about providing some real help from your social media channels? Preferably without requiring a long list of personal details.

Featured image by Biljana Jovanovic from Pixabay.

Failure Demand in action

Posted on By Mark Wilson
This content is 3 years old. I don't routinely update old blog posts as they are only intended to represent a view at a particular point in time. Please be warned that the information here may be out of date.

Recently, my work has involved some analysis of a local authority’s business processes. As part of that I’ve been thinking quite a lot about the concept of “Failure Demand”. For those who are unfamiliar with it, Failure Demand is described by the occupational psychologist and author John Seddon as:

“It is demand caused by a failure to do something or do something right for the customer. Customers come back, making further demands, unnecessarily consuming the organisation’s resources because the service they receive is ineffective. ”

Failure Demand – Vanguard Consulting Ltd

Whilst the Vanguard page is worth a read, there’s another great example of Failure Demand in the “How to break the first rule of Systems Thinking” post from ThinkPurpose.

What does Failure Demand mean in practice?

Any system used to provide a service has a given capacity. To use this efficiently, there is a balance between reducing resources and managing demand.

On the resource side, we can look at how resources are used:

  • Do we have the right people and skills?
  • Are they motivated and focused?
  • Are processes efficient?
  • How is IT used?
  • Can self-service help?

When it comes to demand, the first question to ask is not be how effective the use of resources is. We should really ask are they doing the right thing? Does it meet the customer need?

If it doesn’t then there will be repeat contacts, often relating to Failure Demand – where the volume of work is increased by managing incidents of failure within a process. Examples of Failure Demand include “you’ve sent the wrong item” or “the person didn’t meet the agreed appointment”.

It often takes longer to put something right than to get it right first time. An organisation can implement the very best systems but if it doesn’t meet customer needs in will fail. This is true whether that customer is internal or external; paying for a service or not; client, citizen, traditional “customer” or student. Customers will become frustrated and annoyed that they have repeated contacts to avoid issues. Staff suffer reduced morale as a result of their increased workload.

A real world story of Failure Demand

I spent a good chunk of one day last week working from a car dealership. It doesn’t matter which one… this could have been one of many up and down the country. I also know they are really hot on customer satisfaction. I’d like to make it clear that all of the staff involved were friendly, attentive and did their level best to help me. This is not a complaint, just a true story that helps me explain the Failure Demand concept.

My car is 3 years old, so it was booked in for a service, statutory MOT test, warranty checks, and a quote for an extended warranty.

As the day went on, I saw the Service Manager getting more and more stressed. He wants to do the best he can for his customers but the team is down from 4 to 2 at the moment. That’s going to be tough, but then we layer on the Failure Demand.

At 12:30, my car was nearly ready (it just needed cleaning) and I paid the bill. That was proactive, working to close my account and get me on my way. Great customer service, nothing so far to detract from the outstanding feedback that the dealer hopes to receive (maybe I’ll come back to that in another post).

But I asked about the warranty quote I had requested a week earlier. The person who could deal with that was off work (for understandable personal reasons) but the receptionist who had booked my appointment had assured me it would not be a problem. so, a message was left and someone will call me back after the weekend (Failure Demand 1).

At 13:30, I chased up to see why I was still waiting for my car. It hadn’t been cleaned (Failure Demand 2).

At around 14:00, I got my car back. The service handbook had been stamped and details added for the third service but the second was blank. I always take my car to this dealer, so it must have been missed last year. So the Service Manager looked up the details and added them to the book (Failure Demand 3), once he had found his stamping machine.

By now, I was embarrassed that I kept on going back with “things to fix” and I drove away. As I left, I found that my seat was in the wrong position, the dashboard display was unfamiliar, the doors automatically locked (and much more besides). The profile settings associated with my key were missing!

Having heard the receptionist fielding calls to try and let the Service Department focus on customers who were already in the building, I knew that phoning would not get me an answer any time soon. So, I returned to the dealership to see if the settings were lost for good, or backed up somewhere (Failure Demand 4).

Another Service Manager confirmed that they are not backed up. Some software updates are non-destructive. Others less so. So I left again, disappointed.

Except, as I started the car, my seat moved itself, the dashboard was set up as I expected! My profile had loaded but, presumably the software update had been incomplete before. Now it had finished, everything was back (phew).

Later that day, I received a text message. It contained a link to the video report of the inspection on my car during the service. Nice to have, except I’d authorised the repairs hours previously. Not exactly Failure Demand, but potentially another issue to fix in the process.

So, what’s the answer?

The intention is to move to a position where available system capacity is focused on “Value Demand”. Value Demand is characterised with things that deliver value to the customer or to the organisation, such as provision of information, or just getting it right first time.

If the warranty quote was ready when I paid the bill, the car had been washed when I was told it would be, and the service handbook had been stamped first time then I would have been happy and three items of failure demand could have been avoided. If the Service Manager had known to tell me that software updates might still be taking effect when the car was restarted I might have been less concerned about the missing profile.

The customer would have been happier, the Service Department’s workload would have been lower, and the Service Manager would have been less stressed.

It seems that spotting these issues as a customer is easy… the trick is working out how to fix them in my own work processes…

Featured image: author’s own.