[drop_cap]A[/drop_cap] customer just hit the moment your product was built for. They ran the report that used to take a day of spreadsheet work in ninety seconds. They closed the deal your dashboard helped them see coming. They finally, after three weeks of half-attention, understood what the product actually does for them — and for one window of time, they would tell anyone who asked that it’s worth it.
That window closes. Not because they stop liking the product, but because the feeling fades into the background of a normal Tuesday, same as any tool that just works. Most B2B software companies never notice the window at all — they ask for a review on a fixed schedule (thirty days after signup, or once a quarter, whatever the automation says), which means they’re almost always asking either too early, before the value has landed, or too late, after the feeling has gone flat. Stage 9 is where you stop guessing at the calendar and start watching the product for the moment itself.
Why This Channel Beats the Others
Every other channel you’ve built in this playbook competes for the same finite thing: a stranger’s attention, bought or earned, in a market where every other B2B tool is bidding for the same keywords and writing the same category-benefit headlines. The customer who already pays you doesn’t need convincing. They need only two things: a reason to say something, and a low-friction way to say it. That’s a fundamentally cheaper unit of growth than another ad dollar or another cold outbound sequence, and it compounds — a customer who refers two others, one of whom refers two more, grows the way paid acquisition never can on a flat budget.
The asset stack for this stage is four things, built roughly in this order: the review, because that’s what a buyer checks before they’ll take your call; the case study, because that’s what turns a considered buyer into a confident one; the referral, because a business buyer’s currency for a favor isn’t a coupon; and the product-led share surface, because some of your best distribution happens inside a workflow you already own, if you build it to carry your name honestly.
The Review — Where B2B Buyers Actually Check
Consumer reviews live on Yelp and Google Maps. B2B reviews live somewhere your buyer will actually look before a purchase decision: G2, Capterra, TrustRadius — whichever software-review site your ICP consults, and it’s worth finding out which one that actually is rather than assuming. A prospect evaluating three vendors reads the reviews on the shortlist before they ever take a demo call, and a thin or absent review profile reads as a strike against you, no matter how good the product is behind it.
The honest way to ask sounds almost too plain: tell the customer specifically what a good review does for you (helps other companies like theirs find and trust the product), and tell them exactly what to write about (not “leave a review,” but “what specifically changed for your team since you started using it”). A vague ask produces a vague review — three stars, “it’s fine, does what it says” — which does nothing for a buyer trying to picture their own outcome. A specific ask produces a specific review, and specific reviews are what actually move a purchase decision.
Never pay for a five-star review, and never gate a review request behind a satisfaction score the way a consumer product might. A paid or coerced review is a liability, not an asset — most B2B review platforms have integrity policies that can delist a profile caught gaming it, and a customer who senses they’re being managed toward five stars will write a worse review out of spite, not a better one. Ask everyone in your qualifying group; let the honest ones write what they actually think. A profile of forty reviews averaging 4.6, with a few honest fours naming a real rough edge, reads as more credible to a sophisticated B2B buyer than a suspiciously perfect five.
The Case Study — The Structure That Gets Read
A case study that opens with your product’s feature list gets skimmed and closed. A case study that opens with a situation the reader recognizes gets read to the end, because the reader is looking for themselves in it, not for you.
The structure that works, in order:
- The situation — who the customer is (company size, industry, role of the person you’re quoting) and what wasn’t working before. Not “they needed a better solution” — the actual, specific friction: the report that took a day, the seat count they’d outgrown, the manual process three people touched every week.
- The specific number that moved — the one metric the customer would defend in a room. Time saved per week, revenue attributed, headcount not hired, churn reduced by some percentage over some number of months.
$X saved,X% faster, “your own number” — never someone else’s benchmark standing in for this customer’s result. - The objection they had before buying — this is the section most case studies skip, and it’s the one doing the real work. Every prospect reading the case study has their own version of the doubt this customer had (too expensive for our size, too hard to migrate off the old system, another tool nobody will actually use) — naming it and showing how it resolved does more persuasive work than any feature description.
- What they’d tell someone else considering it — close in the customer’s own words, not yours. A direct quote here carries weight a paraphrase never will.
The Interview That Produces It
A good case study is written from a real conversation, not reverse-engineered from a support ticket. Run this sequence, in order, and let each answer lead into the next rather than working through it as a checklist:
- What was the situation before you started using [product] — what were you doing instead, and what wasn’t working about it?
- What almost stopped you from buying? What was the doubt or objection you had to get past?
- What changed once you were actually using it — walk me through the first real win.
- Is there a number you’d point to — time, money, headcount, anything you can put a figure on?
- If someone in your position was considering this today, what would you tell them?
Record it if the customer is comfortable with that; a direct quote pulled from the actual conversation reads as more credible than one that’s been smoothed into marketing copy. Ask permission to publish before you write a word, and send the draft back to the customer for approval — both because it’s the professional thing to do and because a customer who reviews their own quote before it ships rarely asks for a change, and often adds a better line than the one you had.
The Referral — What to Offer a Business Buyer
Consumer referral programs run on a discount: refer a friend, both of you get 20% off. That currency works when the product is a monthly subscription a person pays for out of their own pocket. It works less well for a business buyer, because the decision-maker referring you isn’t usually the one who’ll feel a discount on next month’s invoice — someone else approves that budget line, or the account is priced per-seat in a way a flat discount barely touches.
What lands better with a B2B referrer:
- Account credit applied directly, rather than a percentage off — a clean, visible line the referrer can point to.
- A public thank-you — a shout-out in your customer community, a co-marketing mention, something that reflects well on the referrer’s own reputation with their peers, which often matters more to a professional than a small discount does.
- Extended access or a feature unlock — if your product has a usage ceiling (seats, API calls, a premium tier), unlocking headroom for the referrer costs you little at your margin and is worth more to them than the equivalent cash.
Whatever you offer, make the ask concrete and time it to the moment described at the top of this chapter — right after the customer has just told you, unprompted, that something worked. “You mentioned this saved your team real time — do you know another [role] who’s dealing with the same thing?” is a real question asked at a real moment, not a banner ad in the product.
The Product-Led Share Surface
Some of your best distribution happens inside the product itself, if you design for it honestly. A report your customer exports and sends to their own client. A dashboard link they share with a colleague. An integration badge that shows up in a tool their team already uses. Each of these is a moment your product’s name travels to someone who has never heard of you, carried by a person they already trust — which is a fundamentally different kind of exposure than an ad they’d scroll past.
The rule that keeps this honest: it has to be genuinely useful to the person sending it, not a marketing insert disguised as a feature. A “Powered by [product]” watermark on an export that the sender didn’t ask for and can’t remove reads as a tax on the user, not a feature — and a savvy B2B buyer notices the difference immediately. A shared link that actually renders well for the recipient, that carries your name because the sender chose to leave it there, is the version that compounds.
Timing the Ask — Watch the Product, Not the Calendar
The single highest-leverage decision in this whole stage is when you ask, and the answer is never “on day 30” or “at renewal.” It’s whatever your own usage data tells you is the moment a customer has just realized value — a first successful report generated, a usage milestone crossed, a support ticket resolved that turns into a five-star reply, an NPS score that just came back high. Any of these is a real, observable signal that beats a fixed schedule, because a fixed schedule fires on customers who haven’t gotten there yet just as often as it fires on customers riding the high of a genuine win.
Build the smallest version of this you can actually observe today: pick one in-product event that reliably correlates with “this customer just felt the value,” and trigger the ask from that event instead of from a date on a calendar. You’ll get it wrong at first — every team does, tuning which event actually predicts a good mood versus a customer who’s simply still logged in. Refine it as you learn, but start from an event, not a date.
When a Customer Says No
Some customers won’t leave a review, won’t accept a referral offer, won’t sit for a case study interview — and that’s a normal outcome, not a failure of the ask. Don’t chase. A second ask after a clear no reads as pressure, and pressure is the fastest way to turn a quietly satisfied customer into an annoyed one. Thank them for their time, note it so you don’t ask again next quarter, and move on to the next customer who might say yes.
The one worth following up on is a soft no — “maybe later,” “let me think about it” — where a single, well-timed reminder weeks later (never sooner) recovers a real share of genuine intentions that simply got lost in a busy week. Beyond that one reminder, let it go. The customers who do say yes are the ones building the asset this stage exists to produce.
AI Earns Its Place Here
This is a stage where a model does real drafting work once you feed it the specifics. Give it the customer’s situation, the number that moved, and the objection they had before buying, and ask it to draft the case study in the four-part structure above — then check every draft against the transcript, because a model will smooth a customer’s actual words into something more polished and, in doing so, less true.
The tutor beside this lesson can also help you design the review-ask message and the referral invite so each one names something specific about this customer’s use of the product, rather than reading like a template every SaaS company sends. Feed it your actual product, your actual customer segment, and the in-product event you’ve chosen as your timing trigger, and ask for a first draft of each message — the review ask, the referral invite, the day-3 follow-up nudge. Treat every draft as a starting point: the judgment of which customers to ask, when the moment is real, and when to simply say thank you and move on stays yours.