Research methods

How to collect customer feedback in B2B: eight methods and when to use each

Eight ways to collect customer feedback in B2B, from support tickets and short surveys to scheduled calls, with what each can tell you and how to choose.

Instant Expert EditorialPublished 6 min read

The practical way to collect customer feedback in B2B is to combine a few passive channels that catch problems as they happen (support tickets, an in-app feedback link, account notes) with a few active ones where you ask on purpose (a short survey at the right moment and scheduled calls). Decide what you need to learn first, then pick the cheapest channel that can answer it. Passive channels tell you where people struggle. Calls tell you why.

B2B has two quirks that shape the choice. You usually have tens or hundreds of customers, not millions, so a survey can produce too few answers to count, while a handful of conversations can cover a real share of your accounts. And the person who uses the product is often not the person who pays for it, so you need to hear from both.

A worked example

Suppose you sell employee onboarding software to 40 mid-sized companies. This example is hypothetical. HR coordinators use it daily; HR directors sign the contract. You want to know why some accounts invite every new hire through the product while others still send paperwork by email. We will come back to this below.

Start with the question you need answered

Write one sentence: "We need to learn ___ so we can decide ___." For the onboarding company: "We need to learn why some HR teams keep using email for paperwork, so we can decide whether to fix the document flow or change onboarding." That sentence rules channels in and out. An NPS score will not answer it. Ticket tags might show where people get stuck. Calls with coordinators at the email-heavy accounts will show what they do instead and why.

Eight ways to collect feedback, and what each is good for

MethodBest forEffortWatch out for
Support tickets and chatWhere people get stuck, in their wordsLow, once taggedOnly hears from people who contact support
In-app feedback linkReactions at the moment of useLowSelf-selected, often short
Short survey after a key eventDirection of sentiment across accountsLow to mediumSmall B2B samples; reasons missing
Account check-insRenewal risk and priorities of the buyerMediumBuyers may not know daily problems
Sales call notesObjections and competing optionsLowSecondhand and summarized
Scheduled customer callsWhy something happens and what it costsMedium to highTakes time to recruit and run
Calls with lost deals and churned accountsWhy people chose something elseHighHard to book; memories fade
Product usage dataWhat people do, as opposed to what they sayMedium to set upShows what, not why

A few notes on the ones that need care:

  • Support tickets. A team at the UK Government Digital Service built a shared set of ticket categories by having three people tag the same 50 tickets and compare, then tagged new tickets as they arrived and reported every two to three months. That turns a support queue into countable feedback.
  • In-app feedback. NN/g describes customer feedback from links, buttons, forms and email as coming from a self-selected sample. Useful for spotting problems; unreliable for judging how common they are.
  • Surveys. Keep them short and tied to a moment, such as the first completed onboarding or 60 days after launch. If you use NPS, Bain's method is the percentage of 9s and 10s minus the percentage of 0 to 6 answers. With 40 customers and perhaps a few users each, expect a small number of responses and treat changes cautiously.
  • Usage data. Behavior complements what people say. NN/g's framework separates "what people say" from "what people do" and notes the two are often quite different. In the example, usage data would show which accounts send documents through the product; it would not show why the others do not.

Schedule calls for the questions that need a why

Scheduled calls are the most useful channel for B2B feedback when a decision is at stake, because you can ask follow-up questions and hear what happened. GOV.UK's interview guidance notes that interviews can run from 30 minutes to 2 hours, that longer sessions give more detail but make recruiting harder, and that you should focus on stories and real examples rather than how things should happen.

For the onboarding company, a first round could be six 30-minute calls: four coordinators at accounts that still email paperwork and two at accounts that do not. Ask each to walk through the last new hire they onboarded, starting from the signed offer. Where did documents go? Who chased signatures? What did they do when something was missing? Comparing the two groups is where the answer usually shows up.

Some practical points:

Keep everything in one place

Feedback scattered across a help desk, a CRM and three people's notes is hard to use. Pick one place, give every item the same few fields (account, role, date, source, tag, quote) and review it monthly. A simple research repository is enough at this size. If you want to turn the channels into a standing program with owners and a reporting rhythm, see our guide to voice of customer programs.

Close the loop

When a change ships because of feedback, tell the people who asked for it. It shows customers that speaking up leads somewhere, which makes the next request for feedback easier.

Your next step

Write your one-sentence question, check which channel can answer it, and book the first three calls this week. If the people you need are not your customers, such as HR coordinators at companies using a competitor, search for people by describing the role, or start from the people and talent professionals in HR technology directory page. Instant Expert finds people whose work matches your question. You review them, it sends the invitations, and you pay for each call that gets booked.