Written by Tim Butler, CEO, Innovation Visual
When leaders choose their first AI project, one instinct leads them wrong almost every time.
They start with admin, data entry, the repetitive junior tasks nobody enjoys. It feels like the safe place to begin, low risk, easy to explain, and nobody complains when the admin gets faster. But in almost every case we have seen, this is the wrong place to look first, and this blog explains why.
Starting with the easiest task to automate answers the wrong question. It optimises for what is simple rather than what is valuable, and those are very different things. The biggest returns almost always sit where scarce judgement is being spent on work that does not need it.
This is not about replacing anyone. It is the opposite. AI's job here is to take the low value, repetitive parts of a senior role off someone's desk so they can spend more time doing what they are good at, and what the business needs most from them.
Look across your organisation and count how many people are genuinely rare in it. Not busy, rare. There is usually one CEO. Often only one or two people who carry the judgement and relationships that make the real difference to revenue. Everyone else can, to some degree, be trained, supported or brought up to speed.
Those one or two people are hardest to replace, and their time is the scarcest resource the business has. Amplifying what they do is worth far more than freeing up someone whose role another hire could cover easily.
When you take low value work off a senior person, whether that is preparing board packs, pulling together summaries, or chasing information that should already be to hand, the gain is not simply the hours saved. The role itself does not shrink, it sharpens.
The value is what that person does with the time you have given back. Free up an hour for a junior team member and they typically move to the next task on their list. Free up an hour for a senior person and they might close a deal, unblock a decision that has been stuck for weeks, or spend real time developing their team instead of firefighting. The time saving looks identical on a spreadsheet. The value it unlocks is much higher.
None of this means junior and admin tasks are not worth automating. They often are, and they are a sensible place to build confidence and get a team comfortable with the change. What follows is a junior example, and it is worth reading with that in mind.
We ran a benchmarking exercise for a business that was manually importing data after every event it ran, roughly once a month. A junior team member was spending around 10 hours on it each time. Automating that step saved around 120 hours a year, worth about £6,000 in salary time. On its own, that is a reasonable case for automation.
But when we looked at what else the automation touched, the picture changed. Because the data was cleaner and followed up faster, conversion improved too, moving from 34% to 38%. That is roughly one extra deal a month, worth around £64,000 in additional revenue a year at the business's average deal value.
The conversion gain was worth more than ten times the time saving everyone had been focused on. Nobody was looking for that number when the project was scoped. It was only revealed because someone asked what else changes when this task disappears, not just how many hours it removes.
This was a junior task. It is worth noticing what that means. If an unglamorous data import concealed a revenue gain worth ten times the hours it saved, the same question asked of a senior person's week has more to find, not less. The lesson is not that junior automation fails to pay back. It is that almost nobody asks the second question, and the answer is usually bigger where the judgement is scarcer.
A few patterns keep appearing in the businesses getting genuine return from AI, rather than just automation for its own sake.
They look at revenue as a series of steps. From finding the right prospects to engaging them, converting them and closing them, they look for where a small improvement compounds rather than where the biggest pile of manual work sits. A 10% improvement compounds because each step feeds the next, better leads convert better, faster follow up closes more deals, and small gains stack rather than sit side by side. Improve each step in that journey by 10% and the gains accumulate to something far larger than any single step on its own.
They look for areas genuinely underperforming. They put focused effort into fixing those areas, rather than automating everything evenly across the business.
They think about how to amplify their strongest people, rather than only freeing up the weakest link. If someone in the business is exceptional at something, the more valuable question is how you multiply what they do.
If you want to work out where your own highest value opportunity is sitting rather than guessing, that is exactly what we help leaders map out in a day. Find out more about the AI Workshop for Leaders.
Or if you would rather talk it through directly, book some time with me using my meeting link.
Here are a few questions that come up whenever this topic gets raised with leadership teams.
No. It is often still a sensible first step, particularly for building confidence in AI across a team. The point is not to assume it is where the biggest financial return sits.
Start by mapping what your most senior or highest value people spend their time on in a normal week, then look for the tasks that are repetitive, low judgement, or purely administrative. Those are usually the first candidates.
Because time saved tends to convert into modest efficiency, while a conversion improvement compounds against existing revenue. A small percentage shift can be worth several multiples of the hours it took to get there, as our example showed.