Start with the ICP your CRM already knows
The best ICP evidence is usually sitting in your closed-won and closed-lost records. Before any workshop, we pull every opportunity from the last 18 to 24 months and look at what the winners share: industry, headcount band, tech stack, the trigger that started the conversation and the title that signed. We then compare that with the deals that stalled after the second meeting.
Most teams find their real ICP is narrower than the one in the pitch deck. That is good news. A narrow ICP makes lists cheaper to build, messages sharper and results faster to read. When KNNX, formerly DLT Labs, started with us there was no funnel at all. Defining who to go after came before a single campaign, and the program went on to build $12M in pipeline over 28 months.
- Firmographic filters an SDR can apply in Sales Navigator or Apollo without interpretation
- Trigger signals such as funding, leadership hires, job posts naming a platform, or a competitor contract ending
- Account tiers: Tier 1 for 1:1 work, Tier 2 for 1:few clusters, Tier 3 for programmatic outreach
Positioning that sales will actually repeat
Positioning fails when it is written by marketing and ignored by sales. We write it from call recordings and customer interviews, then test it where it matters: in cold email subject lines, LinkedIn openers and the first five minutes of discovery. If reps paraphrase it without prompting, it works. If they apologize for it, it goes back to the drawing board.
The document itself is short. One category point of view, the costly problem you remove, three proof points and the one reason to switch now. Everything else, from landing pages to webinar titles, is derived from it.
- Persona-level value props for the economic buyer, the technical evaluator and the end user
- An objection map built from lost-deal notes, with the proof that answers each objection
- An entry offer that lowers perceived risk: assessment, benchmark, pilot scope or workshop
Run channel tests like experiments
A channel test without a hypothesis and a stop date turns into a permanent line item. Each test we run has a defined audience, a message variant, a minimum volume and a decision rule agreed in advance. Outbound to a trigger-based Tier 2 list might need several hundred contacts over six weeks to produce a readable signal. A 1:few ABM cluster might need only 15 accounts and one roundtable.
The choice of channels depends on your market. Crypto and digital asset companies often cannot run paid ads, so organic, events and community carry the load. North American supply chain buyers respond to peer rooms and account clusters. The yard and port management software program we ran used port-cluster ABM and CXO roundtables to build $8M in pipeline in 9 months.
Read results in the right order. Reply and acceptance rates tell you whether the message lands. Meeting rates tell you whether the offer is strong enough to earn time. Meeting to opportunity rates tell you whether the ICP is right. A channel can look healthy on the first number and fail on the third, and that failure is usually a targeting problem, not a channel problem.
- One variable per test: audience, message or offer, never all three at once
- A minimum volume and a decision date set before launch
- A written verdict for every test, including the ones that failed, so the same idea is not quietly retried next year
Pitfalls we see in almost every GTM reset
Three mistakes show up repeatedly. First, measuring leads instead of qualified meetings and opportunities, which rewards volume over fit. Second, changing the ICP, the message and the channel at the same time, so no one can tell which variable moved the result. Third, skipping CRM hygiene, which means pipeline created by marketing is invisible when the board asks.
A pivot makes all three worse. When Quills AI moved toward SI resale priced per instance, the buyer, the channel and the sales motion all changed together. The fix is sequencing: settle the ICP and offer, instrument the funnel, then test one channel variable at a time. It feels slower in the first month. It is almost always faster by the end of the second quarter, because every result can be read and acted on.
- Agree qualification criteria with sales before the first meeting is booked
- Keep one owner for the ICP document and version it when it changes
- Review test results every two weeks, not at the end of the quarter
Go-to-market terms, defined
GTM work produces decisions that every later stage of the funnel depends on. Clear definitions keep founders, sales and marketing aligned on what those decisions mean in practice.
- Go-to-market (GTM) strategy: the plan for who you sell to, what you say, what you offer first and which channels reach those buyers.
- Ideal customer profile (ICP): firmographic, technographic and trigger criteria describing the accounts most likely to buy and succeed.
- Buyer persona: the goals, concerns and evaluation criteria of a role on the buying committee.
- Account tiers: groupings of named accounts by fit and timing, which set how much personalization each receives.
- Positioning: the category you compete in, the problem you solve best and why you win against alternatives.
- Entry offer: a lower-risk first step, such as an assessment, benchmark or pilot, that makes a first meeting easier to accept.
- Channel test: a time-boxed experiment with one variable, a minimum volume and a decision date.
- Test verdict: the written scale, iterate or stop decision at the end of a channel test.
- Objection map: common reasons deals stall or are lost, paired with the proof that answers each.
- GTM motion: the main way deals are created and closed, such as sales-led, product-led or partner-led.
- CRM instrumentation: fields and stages that track each deal from first touch to opportunity.
How to evaluate a GTM partner
Go-to-market is one of the easiest services to sell and one of the hardest to judge, because the output can look finished long before it has been tested with buyers. A strong partner makes decisions you can verify in market within a quarter. Use these checks when comparing agencies, consultants or a fractional leader.
Look for evidence that the partner has taken a company from no funnel to measurable pipeline. KNNX started with no funnel and built $12M of pipeline in 28 months after its GTM, website, ABM and webinars were built from scratch.
- They ask for CRM exports, win and loss notes and customer interviews before proposing an ICP.
- They write ICP criteria that an SDR can apply in Sales Navigator without interpretation.
- Positioning is tested with real buyers, not only approved in an internal workshop.
- Every channel test has success criteria and a decision date agreed before launch.
- They instrument the CRM so pipeline can be traced back to segments and channels.
- The engagement ends in live campaigns and documented verdicts, not only a slide deck.
- They are willing to tell you a channel failed and should stop.
- Everything they build, from account lists to messaging, belongs to you.
