ICP, persona and messaging, because most reply rate problems are targeting problems.

The diagnostic question

Before rewriting a single subject line, we ask which accounts closed fastest, which churned, and which never should have entered the funnel. The pattern in that data is usually a sharper ICP than anything written in a positioning document, and it is the method behind an ICP sales will actually use.

Loose targeting is expensive twice. It wastes outbound capacity upfront, then wastes AE time on meetings that were never going to close.

What the work covers

• Ideal Customer Profile. Firmographic, technographic, and behavioural criteria, with explicit disqualifiers.
• Buyer personas. The economic buyer, the champion, and the blocker, with what each one actually cares about.
• Market and account segmentation. Tiering so the highest value accounts get a different level of effort.
• Messaging framework. Problem led narratives per persona, with proof points mapped to each claim.
• Qualification criteria. The definition of an SQL, written down and enforced in the CRM. See how BANT, MEDDIC and CHAMP compare and how to move MQL to SQL conversion.

Why messaging fails even when it is well written

Good copy about the wrong problem is still ignored. Most B2B messaging describes the product rather than the situation the buyer is in on the day the email arrives. We build messaging from the buyer's problem, then map each claim to a proof point, so what the rep says on a call matches what the sequence promised.

Sharper targeting and qualification is the mechanism behind conversion improvements like moving MQL to SQL rates from 10% to 25%, a career outcome at Locus. The lever was rarely the copy alone.

Why the buying group changes the target

Targeting matters more than most teams think, because you get almost no time to work with. Gartner's B2B buying research found buyers spend only about 17% of the whole purchase deciding process meeting with suppliers, and when they are comparing vendors, as little as 5 to 6% with any one rep (Gartner, 2017 to 2024). Most of the journey happens without you in the room.

It is also not one person. A typical buying group for a considered purchase is six to ten people, each arriving with four or five pieces of information they gathered on their own (Gartner). Forrester puts the stall rate at 86%, usually because one stakeholder's concern never got addressed. So a loose ICP does not just waste sending capacity, it spends your tiny window on the wrong accounts and the wrong people inside them. That is why the personas here cover the whole group, the economic buyer, the champion and the blocker, not only the person who replied. The message has to give the champion something to forward, because they are the one selling internally when you are not there.

What a sharp ICP actually produces

A useful ICP is not a paragraph in a deck. It produces two things a rep can act on the same day: a written disqualifier list, the accounts you will deliberately not chase this quarter, and a tiered, named account list where the best fit accounts get a different level of effort. If the ICP work does not change who a rep prospects on Monday, it has not really been done. Loose targeting stays comfortable because nobody has to say no to a segment; a real ICP forces that choice.

Where has this ICP work been done?

Twice from zero, and once inside an existing team.

For NuVista AI, an AWS Advanced Consulting Partner based in Atlanta offering data platform modernization and AI-powered solutions, with no outbound motion at all, GrowthStack defined the markets, ICPs and personas before a single sequence was built. That ordering is the whole method: the account list is the output of the ICP work, not something assembled afterwards to fill a sequence. The engine that followed generated approximately $50,000 in revenue plus additional qualified opportunities.

At Landmark Group's Logistiq division, Shobhit Gupta created the ICPs and personas alongside the CRM design and the sequences, for a SaaS division selling into the Middle East and India. That structure supported $500K ARR in the first quarter, built through targeted outreach.

For WorkStore, the targeting problem was different and more common. The leads already existed, arriving through Meta and Google Ads, and nobody had defined which of them were worth a call. Applying a written standard moved qualification from 5% to 30% and closed approximately ₹15 lakh within two months.

In all three the deliverable was the same: a named account list with verified contacts and a written disqualifier set, not a positioning document.

References

• Gartner. The B2B Buying Journey, Digital B2B Buyer Survey. 2017, reaffirmed 2023 to 2024.
• Forrester. The State of Business Buying. 2024.

Go deeper

• The same problem, seen from the marketing side
• How to build a B2B ICP sales will actually use
• The B2B sales tool stack, the layer after targeting
• AI as a layer, not a bolt-on tool

Who this is for

• Teams whose reply rates fell without the copy changing. That is usually the list drifting wider, not the writing getting worse.
• Teams selling to more than one persona. One message aimed at three buyers persuades none of them.
• Teams where sales and marketing disagree on a qualified lead. The argument ends when the definition is written down and enforced in the CRM.
• Teams about to scale spend or headcount. Fixing the target before adding volume is cheaper than fixing it after.

Who this is not for

• Your reps already work from a named account list. If the list exists, is current, and reps actually use it rather than searching a tool each morning, the targeting work is done.
• You sell to one obvious buyer with no segmentation problem. Where there is a single persona and a single motion, an ICP rebuild is expensive theatre.

Not sure which applies? Run the outbound readiness scorecard, which takes a couple of minutes.

Before the ICP narrows the list, the list has to be sized. That is TAM, SAM and SOM.

This page describes the engagement. For the method on its own, written so you can run it yourself, see how to build a B2B ICP sales will actually use, and for sizing the market the ICP sits inside, TAM, SAM and SOM.