Almost every conversation we have about marketing ROI in accounting firms starts in the same place. Should we run Meta ads. Is Google better. Has LinkedIn stopped working. These are reasonable questions, and they are also the wrong ones to start with, because the channel is only the delivery mechanism for a decision you made earlier.

That earlier decision is strategy, and in practice it is two things: which businesses you are trying to reach, and what you say to them when you get their attention. Everything downstream, including cost per lead, close rate, average fee and how much of your week gets eaten by low value work, is shaped by those two answers.

What follows is the argument we make with firms before we build anything, along with the failure pattern we see most often.

A channel is a way of putting a message in front of a group of people. It has no opinion about which people, and no opinion about the message. That is why two firms can run the same platform, the same budget and the same agency, and end up with completely different results.

When a campaign fails, the usual response is to change the platform. We rarely see that solve anything, because the money never came back from the channel in the first place. It came back from the match between an audience with a real problem and an offer that answered it clearly. Change the tactic and you keep the same match. Nothing moves.

The version of this that costs the most is not an empty pipeline. It is a full one. Enquiries arrive, calls get booked, proposals go out, work comes in, and at the end of the quarter the numbers look almost identical to the quarter before.

We see it in firms where the diary is genuinely full. Staff are stretched, the owner is doing evening admin, and the recovery rate on the new work sits below the existing book. Volume created activity without creating profit. The uncomfortable part is that nothing in the marketing report shows this. Reports count enquiries, and enquiries were exactly what the system was asked to produce.

It helps to define the word before using it. In our work, strategy means two decisions and nothing else. Who are you talking to, and what are you telling them. Budget, platform, creative format and landing page structure are all implementation.

The first decision narrows the market to a defined group, for example construction subcontractors, veterinary practices or ecommerce sellers crossing a VAT threshold. The second decision states why that group should move from their current accountant to you. Both are business decisions rather than marketing decisions, which is why they cannot be delegated to whoever is running the ad account. Get them right and mediocre execution still returns something. Get them wrong and excellent execution returns the wrong clients faster.

The common approach is simpler than any of this. Spend more, get more enquiries, judge the spend by the count. It is easy to explain to a partner group and easy to put on a dashboard, which is most of why it survives.

It also has a certain logic in a firm that has grown on referrals. Referrals arrive pre-qualified, so a firm used to them has never had to think about who is arriving or why. The natural assumption is that paid enquiries behave the same way and there are simply more of them. Lead count becomes the scoreboard because it is the only number the old growth model ever needed.

A count tells you how many people responded. It says nothing about who they were, what they were willing to pay, or how much of your capacity they will consume. Two hundred enquiries from businesses that want the cheapest possible compliance work and twenty from businesses that need advisory support look the same on the report and look nothing alike in the accounts.

The question the count quietly avoids is the important one. Why should this client choose you rather than the firm down the road or the online filing service. If there is no answer beyond availability and price, the market will supply its own answer, and it will be price. That is the outcome the lead number conceals until the recovery rate exposes it.

Take a common example. A firm needs growth, the phone is quieter than it was, and the fastest available message is a low fee. It works immediately. Price is the one claim every prospect understands without explanation, so cost per lead drops and volume climbs within days.

Nothing about that early data looks like a mistake. The campaign is performing against every metric anyone agreed to watch. Enquiries are up, the cost of each one is down, and the conversion rate is strong because the offer removes the only objection that mattered to the people responding. This is the point at which most firms decide to spend more.

The problem surfaces in delivery rather than acquisition. Clients who chose you on price arrive with disorganised records, expect the fee to cover queries nobody scoped, and question every invoice. Write off time rises. Staff spend their week on work that generates the least margin in the practice.

Then the renewal conversation comes and the same clients shop again, because price was the relationship. The firm now has more clients, more admin, more software seats and roughly the same profit, with less capacity to take on anything better. The leads were easy to get. The money was never there to keep.

The reason this pattern repeats is timing. Marketing decisions report their cost on a delay. The enquiry data arrives in days, the client mix arrives over months, and the effect on margin, capacity and staff turnover arrives over a year or more.

By then the cause is hard to see. The campaign that created the problem may have been switched off long ago and remembered as a success, because on the metrics available at the time it was one. What you are really buying when you approve a marketing message is the composition of your client base in twelve to twenty four months. Very few firms review it that way, which is why the review usually happens after the damage.

The real cost is not a wasted ad budget. Budgets are recoverable. What you cannot recover is the time. A book built on the wrong message takes years to serve, and while you are serving it you are not building the practice you intended to build.

Owners describe this to us in fairly consistent terms. They set out to do advisory work for growing businesses and now run a compliance factory. Their better clients get less attention than their difficult ones. Exit value suffers too, because a buyer pays for a defensible client base and price led clients are the least defensible thing a firm can own. That is the bill, and it is paid in working life rather than pounds.

The fix runs the sequence in reverse. Before any spend is committed, describe the firm you want to be running in two years. Which sectors, what kind of work, what the average client looks like and what the team should be spending its time on.

Once that picture is agreed, the audience follows from it and the message follows from the audience. We use registry and search data at this stage to test whether the niche is large enough and whether demand exists in the places you can reach, so the choice is evidence based rather than aspirational. Only then do channels get chosen, because now there is something specific to say and a defined group to say it to. The tactics become an execution question, which is exactly what they should have been all along.

Firms should know what this trade looks like before they make it, because the early numbers get worse. A specific message to a defined audience produces fewer enquiries at a higher cost per enquiry. Anyone judging the account on lead count will think it is underperforming.

What changes is everything after the enquiry. Prospects arrive understanding what you do and why it applies to them, so conversations are shorter and price is rarely the first question. Work is more repeatable because the client base is similar, which makes onboarding and delivery easier to systemise. Fewer clients, more revenue per client, less time lost to work that never paid. That is where marketing ROI actually comes from.

The practical conclusion is that marketing is a decision about the shape of the business, taken in advance and paid for later. Every message you put into the market is a filter, and whatever it lets through becomes your firm.

So the order matters more than the tooling. Decide what the firm should look like, define who that requires you to reach, work out what would genuinely move those businesses, then buy the ads. Done in that order, the channel question becomes straightforward and largely uninteresting. Done in reverse, no amount of budget or optimisation will produce a firm you wanted.