The CTO Was Always One Slide Ahead

What a digital-strategy presentation taught me about building an argument people can discover for themselves

We'd been in the room about ten minutes when it happened for the first time. My manager had just finished making a point about the limits of connecting individual products in isolation, and before either of us reached for the clicker, the CTO said something close to: so the real question is what you actually do with the data once you can collect it properly. That was the next slide. Not roughly the next slide, the actual point of it.

I remember the small jolt of that, and it wasn't a bad one. We'd spent a long time on the sequencing of this presentation, longer than either of us would probably admit to anyone outside the two of us, and here it was being pre-empted ten minutes in by someone seeing it for the first time. My manager glanced at me, and I could tell we were both thinking roughly the same thing: if he's already there, this is going to go well. Understanding the logic that early usually means a strategy gets adopted rather than merely approved, and that everything political that follows, the budget conversations, the sign-off from other stakeholders, gets easier because the person at the top reasoned his way there himself.

It happened again a few slides later. Then again. By the middle of the deck I'd stopped being surprised and started almost anticipating it myself, watching him lean forward slightly before each new slide the way you can tell someone is about to speak before they do. At one point he was mid-sentence, clearly walking himself through the organisational step that came directly after the one we'd just shown him, and I had to cut in. "That's actually the next slide," I said, probably too quickly. "Give us thirty seconds." He laughed, held up a hand in apology, and let us carry on. It happened enough times over the course of the meeting that by the end I'd genuinely lost count.

What we were presenting was a digital strategy: the ladder of capabilities and organisational changes required to move the business toward a meaningfully higher level of digital maturity. My manager had led most of the framing, I'd built out a lot of the underlying logic and the sequencing, and between us we'd tried to make sure it held together as one argument rather than a set of adjacent ideas. That distinction mattered more than I expected it to when we were putting it together. It's very easy for digital strategy to collapse into a list of things that sound good on their own: connected products, better use of data, new digital services, some AI somewhere in the mix, tighter integration, continuous optimisation once it's all running. Say it like that and it sounds like a menu, and menus invite people to pick whatever looks most appealing rather than what the organisation is actually ready for.

The thing is, none of those capabilities are independent of each other, and I don't think we fully believed that ourselves until we tried to draw the dependencies out on paper. You can't build a meaningful digital service on top of a product that isn't reliably connected. You can't do anything useful with data that isn't being captured consistently in the first place. You can't introduce AI into a process that doesn't yet produce the kind of information AI would actually need. Each step creates the conditions the next one depends on. Skip one and you don't save time, you just attempt the later step without its foundation, and when it fails it looks like the technology didn't work, when really the sequence was wrong from the start.

So the deck wasn't six topics sitting next to each other. It was a chain, deliberately: evidence leading to a conclusion, the conclusion opening an obvious question, the next slide answering it with more evidence, and so on. Less a set of topics than an executable argument, one where each step changes what the audience already believes enough that the next question becomes inevitable rather than optional. I think that's the actual explanation for what kept happening with the CTO. He wasn't free-associating his way to ideas that happened to be relevant. He was reasoning forward along more or less the same path we'd followed while building the thing.

This wasn't relief after a nervous start. We were glad it was happening from that first pre-empted slide onward. What it did prove, though, was something more specific than "he agreed with us." If he'd been wrong about what came next, or drifted somewhere unrelated, that would have told us something had actually failed, that the join between two slides was weaker than we thought and we were leaning on the audience's patience rather than the logic itself. He wasn't wrong, not once in any way that mattered, and every time he got there before us it was one less thing my manager and I would have to fight for once we left that room.

Giving someone a conclusion and asking them to accept it is not the same as giving them evidence in the order that lets them reach the conclusion themselves. The first asks for trust in you. The second asks the logic to stand on its own, and if it does, people end up agreeing because they did the reasoning, not because they were handed the answer. It also isn't manipulation, provided the audience stays genuinely free to challenge a dependency, an assumption, or a conclusion at any point, which he did, more than once. The purpose was never to prevent pushback. It was to make the reasoning visible enough to be pushed back on precisely, rather than in vague terms neither side can quite pin down.

It's also worth admitting that the order in which we arrived at this argument while building it was nothing like the order in which we presented it. There were dead ends, assumptions we only revisited after they caused a problem downstream, dependencies we only understood once we'd tripped over them. None of that belonged in the room, and untangling it from the version we actually presented was, if I'm being honest, most of the real work of the whole exercise. The audience doesn't need our research process. They need the shortest coherent path from the first fact to the last conclusion.

Out of that meeting I've kept something close to a practical test since. For each slide: what does it establish, what conclusion should the audience draw, what question should that naturally raise, and does the next slide answer it? When the answer to the last one is no, either the sequence is wrong or the slide is in the wrong place. Not every presentation needs to be so predictable that the audience could write the next slide themselves. But when an argument genuinely depends on a chain of causes and consequences, the way a digital strategy does, that test is a useful way of checking whether the chain holds, or just looks like it does from where I'm sitting.

What changed is how I judge my own work now, before anyone else sees it. Not whether it sounds convincing read out loud, but whether someone encountering it for the first time could tell me what comes next. If they can, I've probably built something that holds together. If they can't yet, it might still be right, it just hasn't been put in an order that lets anyone check.