Digital strategy connected to revenue

What a digital strategy actually decides

Strategy is a set of decisions about where finite money and attention go, and the decisions are only real if something is being given up. A plan in which every channel gets more effort is not a strategy, it is a wish list, and it usually results in everything being done slightly worse.

Three questions have to be answered. Which channels get the money, in what order, and how will you know within a quarter whether it is working. Anything that does not help answer one of those is background reading.

The output is deliberately dull: a ranked list with numbers attached, owners named, and a defined checkpoint. Dull is the point. A strategy that reads impressively but cannot be acted on next Monday has not been finished.

The audit, and what the numbers usually reveal

We start with everything currently running, including the channels nobody wants to look at, and put actual spend against actual outcome for each.

Three findings recur. The first is a channel absorbing significant budget on the strength of a reputation it established two or three years ago, which nobody has re-examined since. The second is a channel producing genuinely good returns at small scale that has never been funded properly, usually because it is unglamorous. The third is a large volume of activity, most often content and social posting, that is measured by output rather than by outcome and turns out to be contributing very little.

None of this is unusual and none of it reflects badly on the team. It happens because channels get added over time and nothing gets removed, and because the person running a channel is rarely the person best placed to argue for cutting it. An outside review exists to make that comparison in one place.

• True cost per channel, including the internal time nobody bills for
• Outcome rather than output, so activity volume is not mistaken for return
• Where returns fall off as spend increases, which caps sensible investment
• Overlap, where two channels are being credited for the same customers

Attribution, and why the channel reports disagree

Any audit runs into the same obstacle: add up what each platform claims and the total exceeds the number of customers you actually got. This is not fraud, it is every platform claiming credit under its own model, and a customer who saw an ad, searched your name and then opened an email is legitimately claimed three times.

Rather than pretending one model is truth, we work from a hierarchy. Your own sales records settle how many customers there were and what they were worth. Platform data steers decisions within that platform, since that is what the platform optimises against. Cross-channel comparison is treated as directional, and where a decision is large enough to matter, we test it by changing spend and observing the effect on total business rather than on reported attribution.

That last method is unfashionable and reliable. If turning a channel down by half produces no measurable change in enquiries over a full cycle, its reported contribution was not real. We would rather run that test than build a plan on numbers we know are inflated.

The stop-doing list

Every strategy we deliver includes things to stop, and it is usually the most valuable page in the document.

Stopping is harder than starting, because activities acquire defenders. Someone built the channel, someone reports on it, and stopping reads as an admission that the last two years were wasted. Naming that dynamic openly makes the conversation easier, and it is worth saying plainly that a channel being wrong now does not mean it was wrong when it started.

Stopping is also what makes everything else possible. New work with no freed budget and no freed time gets added to a team already at capacity, and it is done badly. A recommendation to invest in something is not complete until it says what gets dropped to fund it.

How a roadmap survives past month two

Most strategy documents are read once and never opened again, because they were written as a finished artefact rather than as a working plan.

The features that keep one alive are simple. Quarterly horizons rather than annual ones, because a twelve-month digital plan is fiction past month four. A named owner for each item, because work without an owner does not happen. A target attached to each item, so the checkpoint has something to check. And a scheduled review at which the plan is explicitly allowed to change.

That last point matters most. A plan that cannot change gets ignored the first time reality contradicts it. A plan reviewed each quarter against actual numbers, with items promoted, demoted or dropped on evidence, stays useful, and the reviews are where most of the value ends up being created.