Fix a pipeline that stalled

Deals are sitting, the number is slipping, and nobody can say exactly why. Diagnose it in order.

4 steps, 16 questions underneath, read for how the best do it

1. 1
   Who you sell to

   Picking a customer worth chasing, and knowing early when a deal is not one.

   How do the best companies keep their ICP sharp as they grow instead of letting it blur into everyone?

   The gist They re-derive it on a schedule from data they already have: which segments renew, which expand, which churn, which cost the most to serve. Then they act on the answer, including the unpopular half, cutting a segment they used to sell to. The best ones also separate the ICP from the total market deliberately, so nobody confuses 'who we could sell to one day' with 'who we chase this quarter'. And they revisit it whenever the product changes shape, because shipping a new capability quietly makes a previously bad-fit segment good, and vice versa.

   📰 Defining our ICP is the most important thing we ever did A real company showing its ICP before and after, including the messy middle where they were guessing. The point that narrow ICPs beat broad ones, a... 4 questions on who you sell to →
2. 2
   Outbound that works

   Cold email, cold calls, LinkedIn, sequences and lists. How to start conversations with people who have never heard of you.

   What are the best outbound teams doing in 2026 that the rest of us are not?

   The gist They treat outbound as an engineering problem, not a headcount problem. The best teams have someone technical sitting inside GTM (the GTM engineer role) who wires together data sources, enrichment and AI to find signals nobody else is acting on, then builds a play around each one. The three rules Clay's own team lives by are worth stealing: find signals that are specific rather than generic, expect your creative edge to decay quickly, and win by iterating faster than the competition. Notice what this is not: it is not more volume, and it is not better templates. It is a tighter loop between a hypothesis and a measured result.

   📄 The GTM Inflection Points That Powered Clay to a $1B+ Valuation Seven specific decisions rather than a success story: the narrow first segment, reverse demos, community-led content, a waitlist held for fifteen m... 4 questions on outbound that works →
3. 3
   Running a sales process and closing

   Discovery, demos, qualification, champions, procurement and the last mile to a signature.

   What separates the top ten percent of AEs from everyone else?

   The gist Discovery discipline, mostly. Top performers spread questions evenly across the whole cycle rather than front-loading a checklist on call one, and they treat discovery as something that continues right up to signature. They uncover three or four business problems instead of one, they name the cost of the status quo in the customer's own numbers, and they bring price up early and confidently rather than hiding from it. They also qualify out faster, which is why their pipelines look smaller and convert better. The gap is rarely charisma, it is that the best reps run a tighter, more curious process and are comfortable losing deals on purpose.

   📄 Effective strategies for successful sales discovery calls Gives you numbers to aim at on your next call instead of vague advice: roughly 11 to 14 questions, spread through the call rather than front-loaded... 4 questions on running a sales process and closing →
4. 4
   Forecasting and GTM metrics

   Pipeline coverage, stage conversion, CAC payback, magic number, and calling a number you can defend to a board.

   How do the very best forecasting teams actually run their week?

   The gist They separate three meetings that most teams collapse into one. A deal review, where the conversation is about strategy on named opportunities and nobody says a number. A forecast submission, where every manager submits a written commit early in the period and it is logged so it can be graded later. And a variance review, where last period's called number is compared against the result and the misses are dissected by cause, not by person. The discipline that makes it work is that the forecast is submitted in writing before the discussion, so the room cannot anchor on the loudest voice. Underneath it all sits the operating cadence: the same meetings, the same order, every week, so the forecast is a routine output rather than a quarterly panic.

   📄 The Cadence: How to Turn Your SaaS Startup into an Army with David Sacks (Video + Transcript) The full explanation of the operating rhythm Sacks ran at Yammer, with the audience Q and A that surfaces the edge cases. Use this version rather t... 4 questions on forecasting and gtm metrics →