How to Tell If Your Marketing Is Working When the Sales Cycle Is Long

Short sales cycle businesses have it easy. Someone's water heater breaks, they search, they call, they buy, all inside two hours. You can point at the ad and the invoice and draw a straight line between them.

Long cycle businesses do not get that. Someone reads a post in February. Downloads something in March. Talks to you in May. Goes quiet. Comes back in September with a budget. By then the ad that started it is buried under thirty other touches, your reporting tool has attributed the sale to whichever link they clicked last, and you are seriously considering cutting the thing that actually worked.

This is the most expensive measurement problem in small business. You cut the channel that's building the pipeline because the payoff hasn't landed yet, and then you wonder why next year is quiet.

Why last-click lies to you

Most analytics tools credit the last thing someone clicked before they converted. That's a design decision made because it's easy, not because it's true.

In a long cycle, the last click is almost always the least interesting one. It's a branded search. They typed your company name into Google because they already knew who you were. Your analytics reports this as "organic search — free!" and gives zero credit to the six months of work that made them type your name.

So the pattern goes like this: your content, your ads, your referral network, your presence — all the slow stuff — build awareness. The awareness eventually produces a branded search. The branded search takes 100% of the credit. And on the report, all the slow stuff looks worthless.

If you make budget decisions on that report, you will systematically defund exactly the things that are working.

Fix one: ask them

The single most valuable measurement upgrade available to a long-cycle small business costs nothing and takes fifteen seconds.

Ask every customer how they heard about you. Not a dropdown on a form — an open text box, or better, an actual question asked by an actual human during the first conversation, and then written down.

"Before we get started — how'd you come across us?"

People will tell you. And what they tell you will not match your analytics, because what they tell you is the story, not the click. They'll say "I've been reading your emails for a while." They'll say "you did my neighbor's driveway." They'll say "I saw you on that podcast last year." Your dashboard cannot see any of that.

The catch: you have to record it somewhere structured, every time, without exception. A field on the job record. A tag on the customer. If it lives in a rep's memory, it doesn't exist.

Do this for ninety days and you will know more about what drives your revenue than any attribution tool has ever told you.

Fix two: measure the middle, not the end

If the outcome is six months away, you cannot manage on the outcome. You'll be flying blind for two quarters and then correcting too late.

So measure the leading indicators — the steps that reliably precede a sale. In most long-cycle businesses the chain looks something like:

Awareness → inquiry → conversation → proposal → close

You can't see closes yet. You can see the rest of it right now.

• How many new inquiries this month, and from where?
• How many turned into a real conversation?
• How many conversations produced a proposal?
• What's the total dollar value sitting in the proposal stage?

If proposals sent are up 40% this quarter, you don't need to wait six months to know your marketing is working. You already know. The revenue is a formality that arrives later.

This is the whole trick. Pick the earliest step in your chain that reliably predicts a close, and manage on that. It converts a six-month feedback loop into a two-week one, and a two-week feedback loop is one you can actually steer with.

Fix three: cohort by when they arrived, not when they closed

Here's a mistake that quietly wrecks long-cycle reporting. You look at "revenue in June" and try to figure out what marketing caused it. But June's revenue came from leads that arrived in December, January, and February. You're grading last winter's work with this summer's report, and matching them up wrong.

Instead, group leads by the month they first showed up. That's a cohort. Then track that cohort forward: of the 60 leads that arrived in January, how many had closed by month three? Month six? What did they end up being worth?

Now you can compare. The January cohort converted at 12% by month six. The April cohort is at 9% at month three, which is ahead of where January was at month three. Your April marketing is working better, and you know it in July instead of October.

This is not complicated. It's a spreadsheet with a row per month and a column per month-since-arrival. It's the single most clarifying thing you can build for a long-cycle business, and it takes an afternoon.

Fix four: track first touch, not just last

If you're going to have attribution data at all, capture the first touch and hold onto it.

The mechanics: when someone first lands on your site, record where they came from and keep it. When they eventually fill out a form — maybe on their fifth visit, weeks later — that original source travels with them. Now you can see both ends of the story: what got them in the door, and what they clicked right before they raised their hand.

Neither is the whole truth. Together they're a lot closer than last-click alone.

Fix five: hold something still

The honest, unpopular answer to attribution is that you will never perfectly know. Human buying decisions aren't traceable. Someone saw you three times and heard your name from a friend and can't reconstruct it themselves.

So run experiments instead of chasing perfect measurement.

Turn one channel off for sixty days. Not forever — sixty days. Watch what happens to total inquiries, not the channel's own numbers. If total inquiries don't move, the channel wasn't producing anything you weren't getting anyway. If they sag, you just learned something no dashboard would have told you.

The discipline: change one thing at a time, and give it long enough for a long cycle to reveal itself. If your cycle is four months, a two-week test tells you nothing.

What this looks like in practice

You end up with a small set of things that fit on one page:

• New inquiries this month, by source, with the source captured at first touch
• The self-reported "how'd you hear about us," recorded on every deal
• Proposal count and total proposal value in the pipeline
• A cohort table showing how each month's leads have converted so far
• Revenue, which you look at knowing it's a lagging echo of work you did months ago

None of that requires software you don't have. It requires that you capture the source at the start, ask the question out loud, and stop reading the last-click report as if it were the truth.

The businesses that win long-cycle games are the ones that can tell the difference between "this isn't working" and "this hasn't paid off yet." That difference is entirely a measurement problem, and it's solvable.

If you want a straight answer about which of your marketing is actually building pipeline, that's the kind of thing I untangle. Get in touch and bring whatever numbers you already have.