B2B Go-To-Market Playbook for SaaS and Subscription Businesses: One Motion per Segment, Tested Against the Pricing Unit

In my view, most SaaS plans stall for one of two reasons, and both hide in blended numbers.

One is the segment. A company ends up selling to enterprises with long cycles and to younger firms that buy in weeks. Averaged, the numbers describe no real customer.

The other is the pricing unit. A seat, an interaction and a hybrid of the two reward different results. When a support team's AI agents take more of the work, the buyer needs fewer seats. On a seat contract, the better the product does, the stronger the case for a smaller renewal. The test works with or without AI: when the customer succeeds, which billed unit shrinks?

Retention comes before the plan. At a GTM teardown with AI founders on 28 March 2026, one company had NRR under 100% and wanted more pipeline. My answer was "Fix retention before scaling acquisition." (LinkedIn, 30 Mar 2026, founder unnamed.)

Guide 01 scores motions per segment, guide 04 checks fit before paid demand, and guide 06 works accounts one by one.

My record: Happay, Vice President Marketing, Nov 2019 to Jan 2022: 161% ARR growth, +178% enterprise deal size, 65 to 72% marketing-sourced pipeline.

Your closed deals, your billing data and a spreadsheet do most of the work. The tools save drafting, and each result is a first draft you check. Each works from what you type, so the split is your job.

One rule runs across the whole method, and it is my suggested rule, not a tool's: no segment gets growth money on a fit score. It gets money on its own gross retention, net retention and payback.

About this example. Kapture CX is a real company. Kapture CX has not reviewed this guide. Facts about Kapture CX come from public sources, linked alongside with the date I checked them. Every number marked illustrative, and every segment, customer type, seat count and churn reason in the example, is an input I chose to show the method. None of it is Kapture CX's data, and the outputs are exercise results, not Kapture CX's real strategy.

What is public: Kapture CX is based in Bengaluru and sells customer experience software with AI agents. Inc42 reported a $10 million pre-Series B round on 30 June 2026. Its pricing page says "Enterprises can choose agent-based, interaction-based, or hybrid pricing", so the seat trap below is a general test, not a view of its contracts. Its site lists BFSI and Digital Native among the industries it serves.

I use those two industries as illustrative segments: "regulated" means banks, NBFCs and insurers with 200 to 2,000 contact centre agents, and "digital native" means digital native consumer brands with 50 to 300 support agents. Nothing here says how Kapture CX sells or what its numbers are.

Step 1. Split the deals, then choose a motion per segment

Start with last year's closed deals. Sort them by who signed, how they found you, cycle and ACV. If the groups differ on those, they need different motions. Then cut the same deals by pricing basis and by how much of the work AI can take, because industry sorts the sales cycle, not seat exposure.

For a first draft, input to the EPIC audit: company_name Kapture CX, industry "B2B SaaS, customer experience software", business_stage series-a, and illustrative self_serve false, geography global, plus the illustrative figures alongside. The regulated challenge reads "Customer experience and support software with AI agents, sold to the contact centres of banks, NBFCs and insurers, moving into new international markets." The digital native one reads "Customer support software with AI agents, sold to digital native consumer brands with small support teams."

What came back: Ecosystem and ABM leads regulated at 9, from "Series A starting point 5; ACV above 50,000 US dollars (E +2); Deal cycle above 90 days (E +2)." Inbound and Outbound leads digital native at 7, and its NRR of 92 adds the line "NRR below 100 percent (leaky bucket: fix retention before scaling acquisition)".

How to read it: the audit starts each motion from a table set by the funding stage you pick, then adds points for lines in your figures. It did not read how these buyers buy. I chose $60,000 and 120 days above its $50,000 and 90-day lines, and $18,000 and 45 days below them, so the different leads come from my inputs. The blended run picks Inbound and Outbound at 8, which leaves the regulated side without an account-based plan. That shows the risk of an average, not proof. The reason to split is that buyers, cycles and proof differ.

Type the 2,400 accounts that match the ICP in step 2, not all 12,000, and the digital native lead moves to Community-Led at 6. So check that motion against how your last 20 deals arrived.

Step 2. Size each segment

Sizing is a count times a contract value, so write down where each count comes from. Input to the calculator, all illustrative: the counts and ACVs from step 1, an ICP match of 25% and a year one share of 2% for regulated, and 20% and 3% for digital native. company Kapture CX, product_category "customer experience and support software with AI agents", data_sources "illustrative count for the exercise", and each segment name as defined above.

What came back: the figures alongside. The calculator multiplies what you give it. Its pipeline is "SOM × 3, a 1 in 3 win rate", marked as an example, so use your own rate. The question is whether 72 small deals is a better year than 30 large ones. That depends on who stays and what each dollar returns, so step 3 follows.

Step 3. Read retention and payback, then the score

Build three numbers per segment from billing data. Logo retention counts customers. Gross revenue retention, GRR, is opening ARR minus churn and contraction, over opening ARR. It cannot pass 100%, so upsell cannot flatter it. NRR adds expansion back. A seat cut leaves logo retention alone, hits GRR at once, and can vanish inside NRR when another account buys more. The bridge alongside is illustrative. Lost accounts here run about half the average size, which is why 26% of digital native logos is 13 points of revenue.

Then payback, at an assumed 72% margin: $45,000 CAC over $43,200 a year is 12.5 months, and $9,000 over $12,960 is 8.3. Regulated is worth more per customer. Per dollar of CAC, digital native leads for six years on NRR, and past year ten on GRR.

The PMF scorecard is a missing-data check here, and no line of the plan rests on it. Input for digital native, all illustrative: product "Kapture CX, customer experience and support software with AI agents, for digital native consumer brands", your_sector saas, business_model saas, time_in_market 2_plus_years, current_metrics "MRR: $90K, ACV: $18,000, Churn: 2.5%, NPS: 31, CAC: $9,000, LTV: $43,000, Retention: 74%, Activation: 71%". LTV is ACV times the 72% margin, over twelve months of churn. Retention is logo retention, though the tool labels that row Revenue Retention.

What came back: on target_market smb_saas, 7, "Emerging PMF"; on enterprise_saas, 5.8, "Searching for PMF". So compare segments only on one band set. The first run scored churn 2.5% as 9, "EXCELLENT", and retention 74% as 3, "CRITICAL", yet 2.5% a month compounds to about 74% a year. An average can outvote one bad row, so the retention numbers above decide. Its own caution is "a rule score from the figures you supplied, not proof of product-market fit". Its priority actions start with NPS, so put retention first yourself.

This is the point where I would want a second pair of eyes: gross retention is below the gate while net retention looks sound. Bring me your segment cohorts, expansion split, CAC payback and proposed spend, and we will work it through in 45 minutes.

Step 4. Test each segment against how you charge

Write what you bill on, and what a customer can cut, and when. Then model the cut. My arithmetic, illustrative prices, not Kapture CX's: take the $60,000 average account as 100 seats at $600, because most accounts start on a few queues, not the whole floor. Suppose AI agents take enough work that the buyer cuts 25 seats. The renewal is $45,000. AI takes short, easy contacts first, so a quarter of contacts is rarely a quarter of seats. Treat this as a scenario and use your own ratio.

Read it in the bridge. If accounts holding 32% of ARR each cut a quarter, GRR falls 8 points to 84 and NRR to 100, while logo retention stays at 89%. Sell 8 points of AI add-ons elsewhere and NRR is back at 108 with GRR still at 84. So watch GRR beside NRR. The company's page says pricing is shaped by "AI features you add on", so add-ons are a real offset, but they show only in NRR.

Next, the rate. Those 25 seats handled about 150,000 contacts a year, at an illustrative 6,000 each. So $0.10 per AI-resolved contact recovers the revenue, and $0.072 plus your run cost keeps the gross profit. That is a floor, not a price. On the buyer's side, at an illustrative $6,000 a year per agent, the 25 seats cost $150,000, about $1.00 a contact. Your price sits between the two. At an illustrative $0.20 you take a fifth of the saving and earn $30,000, twice what the seats paid, and at an illustrative run cost of $0.06 the margin is 70%. Count retries and failed attempts in run cost. For scale, list prices in other markets sit far higher: Intercom lists Fin from $0.99 per outcome, and Zendesk lists pay-as-you-go automated resolutions at US$2.00.

Then test the ranking. Use the simple lifetime sum: ACV times margin, over a year's losses. Seats shrinking 8% a year on top of twelve months of churn take regulated from 8.0x to 4.8x, level with digital native. If both shrink 8%, regulated still leads, 4.8x to 3.8x. So the ranking turns only if regulated automates faster, and its AI-resolved share is the number to watch.

For a first draft, input to the retention playbook for regulated, all illustrative: customer_segment "banks, NBFCs and insurers with 200 to 2,000 contact centre agents"; current_churn_rate "1% monthly"; cs_team_size medium_4_10; product Kapture CX; industry fintech; your_sector saas; churn_reasons "the buyer cuts agent seats at renewal because AI agents do more of the work; the head of customer service leaves and the next one reopens the vendor choice; AI agents stay live on one queue and never spread". Run it as saas_subscription and as usage_based; the schema has no hybrid option.

What came back: the business model changes the health score, led by feature adoption at 30% on seats and usage volume at 40% on usage, both illustrations. The usage run adds a "Volume review" every month and a "Fall alert". That run does not model a seat cut or a price change, so the arithmetic above and the answers alongside are yours. Rewrite its "You told us" drafts, which quote your internal note.

Digital native loses logos more than seats. Reaching 85% logo retention means nearly halving monthly churn, to about 1.35%. My route: the first AI agent live in the first month, a value review before renewal, and fewer seats offered before you lose the logo, which costs GRR, so track both.

Step 5. Put each segment on one page

My call on these illustrative numbers, not the tools'. Take an illustrative $1.6M. Regulated gets $1.35M, the CAC for its 30 customers. Digital native gets $250,000: $180,000 for 20 customers at $9,000, about $360,000 of ARR, inbound only, and $70,000 for the retention route in step 4.

Step 3 says digital native pays back faster, so this split must argue against that. Its lead holds only at $9,000 CAC, and inbound cannot be bought on demand: at about $11,000, 23% higher, its five-year lead on GRR is gone. It keeps 74% of logos and 82% of revenue, so more spend buys more of what leaves, and the teardown call above applies. And regulated is the base that expands. The case against: if digital native's CAC holds at volume and it clears its gate, the money should move. A pre-Series B company might also fund one segment alone, and this plan comes close.

What the regulated money buys, in my view: a named list of about 300 of the 1,500 accounts in SAM, for the 90 opportunities that 30 wins need at the illustrative 1 in 3. That asks 30% of a named list to open within a year, which is a lot, so use your own rates and expect a longer list or a lower target. Start in markets you already sell in, and add one new region only while the gate holds. I hold CAC at $45,000, which a new market may not allow. At month 6, about 45 opportunities are due. Under 25, my suggested line, hold the next quarter's money.

Partners: integrators first. An outsourcer billing per agent loses revenue when AI resolves contacts, so agree how it earns before you rely on it. Have a vendor risk pack ready at the first meeting. My own reference for a named list is Seclore, where I was Director of Product Marketing from July 2016 to October 2017: Rs 2.35 Cr TCV in 6 months through account-based marketing, and 225+ sales meetings across its target accounts.

The gates are retention, not a score, and the marks are my suggested lines, not benchmarks. The illustrative regulated base passes today. In the seat-cut case it fails at GRR 84 while NRR 100 still passes, which is why the gate needs both.

The company already lists hybrid, so the hard part is timing. Write the meter into the first contract on new deals. Move existing accounts one renewal before their AI-resolved share rises, not when procurement asks for fewer seats.

1. Split last year's closed deals into two or three segments, and write ACV, cycle and CAC for each.
2. Choose a motion per segment from how those deals arrived, then run the EPIC audit as a first draft.
3. Size each segment, and replace the 1 in 3 win rate with your own.
4. Build logo, gross and net retention per segment from billing data, then payback.
5. Write what a customer can cut, and when, then model the renewal if AI takes a quarter of the seats.

1. Choosing a motion on blended numbers. The average describes no real buyer.
2. Letting a fit score decide the budget, or comparing scores across different bands.
3. Watching NRR alone. A seat cut hits GRR first, and upsell elsewhere can hide it in NRR.
4. Funding on value per customer without checking return per dollar and payback.
5. Pricing on seats while selling AI that reduces them, or treating the break-even rate as the price. This is a general trap, not a claim about any company.

• How I Check Product-Market Fit, Churn and ICP Gaps Before Scaling Paid Demand Generation
• When High Adoption Hides Churn Risk: A GUARD Health Check, Save Plan, Expansion Case and QBR on One Book of Accounts
• How Much Sales Should a Product-Led Company Add? Scoring EPIC Motions Per Segment, Then Sizing and Anchoring Them
• How to Design a B2B Partner and Channel Programme That Sources Pipeline
• B2B Go-To-Market Playbook for Retail and CPG Sales Tech: Pick the Motion by Brand Tier, Sell to Sales and Finance
• All 24 guides