Most failed engagements were scoped wrong, not executed badly
The instinct when someone describes a growth problem is to start solving it in the room. It feels helpful, it demonstrates competence, and it's how you lose the next three months.
Because the problem as first described is almost never the actual problem. "We need more leads" turns out to be a qualification issue. "Our conversion rate is bad" turns out to be a traffic quality issue. "We need to fix our funnel" turns out to be that sales and marketing haven't agreed on what a lead is.
If you scope against the stated problem, you deliver something competent that doesn't move anything. So before I propose anything, I run the same five questions.
1. What changed recently?
Nobody looks for outside help when things are steady. Something moved — a target went up, a channel stopped working, a round closed, a competitor launched, someone senior left.
The answer tells you two things. First, the actual trigger, which is often not what's in the brief. Second, the political context, which determines whether recommendations can survive contact with the organisation.
If a founder just missed a board target, a twelve-month rebuild of the demand engine is the wrong proposal regardless of whether it's correct.
2. Where does the money actually come from today?
I ask for the last 12 months of closed-won deals with source, deal size, and sales cycle length.
This single request does more work than the rest of the discovery combined. It shows whether they can produce the data at all — which tells you a great deal about the maturity of their operation. And when they can, it usually contradicts the story they've been telling.
The most common finding: 60–70% of revenue comes from a channel nobody is investing in, usually referrals or founder network, while the budget sits on a channel that produces volume and no closes.
3. Who decides, and who can block?
An engagement that reports to the founder and one that reports to a marketing manager with a sceptical VP above them are different jobs with different odds.
I want to know who signs off, who has to implement, and who is quietly opposed. That last one matters most. Growth work almost always requires someone else to change how they operate — sales adopting a new lead definition, product instrumenting an activation event, finance releasing budget differently. If that person isn't in the conversation, the recommendations will be accepted politely and never implemented.
4. What have you already tried?
Two reasons. The obvious one is not repeating failed work.
The more useful one: how they describe past attempts tells you how they evaluate evidence. "We tried LinkedIn ads and they didn't work" is a different organisation from "We ran LinkedIn for six weeks at £4k, got 30 MQLs at £130, and closed one at £12k ACV — the maths worked but we couldn't sustain the creative."
The first will abandon anything that doesn't produce results immediately. The second can be worked with. This changes what I propose and how I stage it.
5. What does success look like in 90 days, in a number?
If they can't answer, we're not ready to scope.
Not "improve pipeline." A number, a metric, and a date. The conversation that produces it is frequently the most valuable part of the whole discovery, because it forces a choice between things that had all been called priorities.
It also protects both sides. An engagement without an agreed number gets evaluated at the end on whatever the client happens to be worried about that month.
What makes me decline
Three patterns, learned the hard way in my first year.
No access to the data. If I can't see closed-won by source, I'm guessing with confidence, and that's worse than not helping.
The real problem is outside marketing. If the product doesn't retain, or pricing is misaligned with the segment they're targeting, demand generation makes the problem more expensive rather than better. Sometimes the honest answer is "this isn't a growth marketing engagement."
They want validation, not analysis. Occasionally the decision is already made and the engagement exists to justify it internally. You can usually tell by question four — every past attempt is described as having been undermined by someone else.
The proposal writes itself
Once these five are answered, scoping is mechanical. You know the trigger, the real revenue engine, who has to agree, what evidence they respond to, and what number you're being judged on.
The engagement I propose is usually narrower than what was originally asked for, shorter, and aimed at the constraint rather than the symptom. That's a harder sell in the room than a comprehensive programme — and it's the reason the work tends to hold up afterwards.
Hilal Tasdan
B2B SaaS Growth Marketing Consultant & Fractional CMO. Partner in Growth.