The reality I have seen…

Alarmbell articles about over-investing in AI with no evidence of ROI are starting to emerge. https://www.youtube.com/watch?v=dx-tMK7w5g8&t=509s. Billions have been spent on this new tech, and we have yet to learn how they will create value. Many promises have been made, some sounding like magic, but most are more assumptions than evidence.

Like many other times, new technology has great potential if applied in the proper context. Over 20 years and at least five cycles of new technology hype, the human context has been the most under-measured and overlooked part of success and investment in technology and IT. Context comes in two aspects: Market or internal fit (does it solve a problem in the right way? Will it not create more significant issues when implemented/) and adoption (Do we communicate it in the right way so that people understand what they are getting and adopt it?). AI has been advertised as the tech that doesn’t need people anymore, so companies have started eradicating the context they need to create and understand value. This approach sounds as risky as crypto-schemes and as challenging to eliminate.

Using context understanding to improve investment and change only requires a few key elements, which are cheap and fast to establish and likely will already exist to some extent.

What are you missing?

The fastest speedboat will get nowhere when placed in a pond. 98% of startups fail, thinking people will come if they have the features. The myth of ‘The feature is so good, how would you not use it?’ should have been buried long ago. It is still part of nearly every business proposition, transformation, change or product launch that I have seen in the last 20 years. You have seen this in various shapes and forms if you run a company or are part of a team.

• Businesses set aside a random budget for AI, blockchain, or influencers because they will undoubtedly lead to new revenue. Yet, people must develop a scenario based on what employees or customers ask.
• Business asks IT to reduce costs; IT buys a cheap, out-of-the-box enterprise platform that gives everyone in the company the same cookie-cutter-reduced set of features.
• Some claim that a supplier’s solution is the most suitable for a company because they know what they are doing or are ‘Industry standard’, so it must be high-quality enough.
• Sales tells IT what to buy based on sales research. It describes how to expand the technology infrastructure effectively in a different way.

Although all these examples have significant flaws, they contribute to about 90% of all transformation or growth efforts.

To understand the significance of the risk, you might ask yourself: “If everyone is using Slack or Microsoft Office, why do some companies end up efficient and others messy and mismanaged?”

“If your IT managers are so smart in picking the right requirements for a tool, why does it create ShadowIT, bad data and rejection by the teams who are supposed to use it to the point where there is damage rather than benefits?”.

This can easily be avoided.

I have seen warehousing solutions that no one used because the data was flawed and the network quality wasn’t considered. I had to code whole Wi-Fi to no Wi-Fi flexible solutions twenty years ago. Today, companies like SAP need to meet these general global requirements.

I have seen IT services expecting employees to hand in their laptops mid-pitch and trust IT that no project data would be lost.

I have seen banking clients use pre-COVID insights on customers to develop post-COVID services, which fell apart on superficial inspection, meaning the bank wasted six months of work.

We know that companies are not just one solid thing. This is why investing 1% in contextual discovery can often result in a 50–90% return.

A typical technology cycle is less than three or sometimes two years, which means looking away for a year without tracking what changes will make your project fail.

To the above points, lack of context discovery leads to:

• Random new-tech budgets fail because how technology creates value is separate from the discovery phase. You then get the ‘newspaper-is-now-digital’ effect instead of the ‘citizen-journalism’ decentralised approach that would re-invent a process through technology and shift the business.
• Reduced tool features are not based on team (sales) needs. While saving money in the short term on IT, they negatively impact sales and what they can do for clients and customers.
• Suppliers need to do the core minimum to maintain their margins. Suppliers are not in the business of an equal growth partnership. Few are commoditised enough to serve as cost-reduction purely. If new technology is a growth factor, you should not outsource it. Response time and level of commitment do not contribute to a supplier’s margins.
• Business and IT have different narratives and roadmaps. One’s target is often counter to the other’s. Leadership has defined different directions for them. Individually or siloed, they make sense. Systemically, they destroy most investments.

Your company is not a factory anymore!

Most products and services are not created in a factory anymore. They all connect with reality. This means that what people expect changes after the product leaves ‘the factory’. Progressive and successful companies have processes that can iterate and improve products to find the market fit they need to create a return on investment. This is exponentially more true for AI. AI only exists with people. People are training and defining the AI systems. People are shaping AI services, and there are, of course, people using AI-enabled services who evolve when using the service. People everywhere. People are your ultimate context.

Go Beyond Tech Requirements!

The benefits that technology can create when considering human needs are massive. Duplicating outdated processes digitally without reviewing them means the ability to turn four-week processes into one-hour ones is seriously repressed.

The government tried to digitise the death certificate service. It deconstructed the previously existing paper form, confusing the whole community, who had to now deal with a service that was detrimental to their efficiency.

Another governmental service digitised paper vouchers. The digital card replacement made a large number of previous receivers of the vouchers being rejected. It generally degrades the service’s success without ever addressing the reason for the change, which was removing the stigma from low-income families to give paper vouchers in exchange for a digital card for the service.

Teams will surprise you!

Change or growth occurs only with the people who work the hours. If you let teams explore and incorporate their insights into the business case, you will at least see a massive level of de-risking.

You can use it on AI or any new massive tool you want to buy and force it on everyone.

You wouldn’t walk into someone else’s party and start playing your music, would you? How will your teams react when you push things on them? You might think you can replace or outsource, but let’s be honest: the limitations of outsourcing contracts and hiring new people every few quarters constitute a significant loss on your bottom line.

Summary

AI needs all the voices. Most technologies today do because they are systemic, respond to user behaviour, and are considered wicked problems. New technology is the most fragile flower that will turn into a flesh-eater if you don’t take care of it, or it will just die. I have seen the death of millions in investment in this context. I have seen companies spending hundreds of millions buying enterprise platforms, and the end users just refused to clean up the data, stopping the project in its tracks.

The Wicked Company is a trusted advisor, non-exec and offers consultant service to companies across a variety of industries.
Say Hello!