Account Expansion & Upsell Research: Finding Your Next Revenue in Your Existing Customers
How to run expansion and upsell research with customer interviews — uncover unmet needs in your best accounts, spot expansion signals before renewal, and build offers customers actually want to buy.
Your cheapest growth is already paying you. Expansion revenue — upsells, cross-sells, and seat growth inside existing accounts — converts at several times the rate of new-logo acquisition and costs a fraction to win. Yet most teams expand by guesswork: they push the next tier and hope. The teams that compound net revenue retention do something different — they research expansion. They interview their best customers to learn what bigger problem the account is trying to solve next, then build the offer around that. This guide shows you how to run expansion and upsell research systematically, and how to make it continuous instead of a once-a-year renewal scramble.
Why expansion is a research problem, not a sales problem
Expansion fails when it is framed as "how do we get this customer to buy more?" That question puts the seller's roadmap first. The customers who expand do so because their situation changed — a new initiative, a growing team, a process they want to standardize, a pain that crossed a threshold. Expansion research finds those triggers before your competitors (or your customer's own build-it-internally instinct) do.
The classic signals — usage growth, feature adoption, support tickets — tell you that something is happening but never why. A spike in usage could mean the account is thriving and ready for more, or that one power user is straining against a limit while everyone else has churned in spirit. You cannot tell from a dashboard. You have to ask.
This is where conversational research changes the economics. Historically, interviewing dozens of accounts about their evolving needs meant a researcher's quarter. With tools like Koji, an AI interviewer runs each conversation — asking your expansion questions, probing every answer in real time, and synthesizing the results automatically. A customer success team can field an expansion study across 80 accounts in a week and walk into renewal season with a ranked list of who is ready to grow and exactly what they want.
What expansion research uncovers
A well-designed expansion study answers five things:
- Adjacent jobs. What related problem does the customer now wish your product also solved? These adjacencies are your highest-fit cross-sell and your roadmap's best signal.
- Expansion triggers. What recent change (headcount, a new initiative, a leadership mandate) created appetite for more? Triggers tell CSMs when to reach out, not just whom.
- Value realized. Where has the product already paid off? Customers who can articulate ROI are your warmest expansion targets and your best case studies.
- Blockers to growth. What is stopping them from rolling you out wider — budget cycles, an internal champion gap, a missing capability, a security review?
- Willingness and packaging. Which higher-tier capabilities feel worth paying for, and how should they be bundled to feel like an obvious upgrade rather than a tax?
Designing the interview with structured questions
Koji blends open conversation with six structured question types — open_ended, scale, single_choice, multiple_choice, ranking, and yes_no — so you capture aggregatable signal and the story behind it in one session. (The structured questions guide breaks down how each type renders in your report.)
An expansion interview plan might run:
- Open-ended discovery: "What is the biggest thing your team is trying to accomplish this quarter?" Probe deeply (2–3 follow-ups). The adjacency you are hunting for usually surfaces here, unprompted.
- Scale: "How much of the value you hoped for have you gotten from us so far, 0–10?" Anchor it so the AI asks what would move the score up — that answer is your expansion thesis.
- Multiple_choice: Present a set of adjacent capabilities and ask which they would use this quarter. Frequency across accounts reveals your strongest cross-sell.
- Ranking: Have customers rank a shortlist of potential add-ons by how much they would value them. This is your packaging hierarchy.
- Single_choice: "Who would need to approve adding budget for this?" — a clean read on the buying committee and deal friction.
- Yes_no: "Would you be open to a conversation about rolling this out to more of your team?" — a warm-lead flag your CSMs can action the same day.
Because every question carries a stable ID from brief through report, Koji aggregates the scale and choice answers into charts and auto-codes the open-ended responses into themes — so the output is "41% of accounts named workflow X as their next priority, concentrated in 50+ seat customers," not a pile of transcripts.
Recruiting accounts without burning goodwill
Expansion research targets your best, busiest customers — exactly the people who will not schedule a 45-minute call. That constraint is why most teams never do it. Koji removes the scheduling problem: a CSM shares one personalized interview link, and the customer completes a voice or text interview whenever it suits them. Voice captures enthusiasm and hesitation; text gets higher completion from executives. Koji's quality gate ensures only substantive conversations (scoring 3+) count toward your plan, so a quick "looks good, no notes" reply never costs you a credit or muddies your analysis.
Interview 15–30 accounts per segment — by tier, industry, and seat count. Expansion appetite clusters by segment, so themes saturate fast.
Turning insight into expansion revenue
Koji's real-time report does the synthesis automatically: ranked adjacent jobs, an expansion-readiness signal per segment, value-realization themes with verbatim quotes, and the blockers standing in the way. Put it to work:
- Build an expansion-ready list. Accounts that scored high on value realized and named a clear adjacent job are your warmest pipeline. Hand them to sales with the customer's own words attached.
- Package around demand, not your org chart. Bundle the add-ons customers ranked highest together; price the upgrade against the value they articulated.
- Arm CSMs with triggers. The recent changes customers described are reusable signals — "just hired a team," "new compliance mandate" — that tell CSMs when to open an expansion conversation.
- Feed the roadmap. Adjacent jobs that recur across accounts are your highest-confidence build bets, validated by paying customers.
The teams with the best net revenue retention do not run expansion research once. They make it continuous — an always-on pulse on their accounts' evolving needs — so every renewal is a known quantity and every upsell is something the customer already told you they wanted.
Common expansion-research mistakes
Three errors quietly kill expansion research. The first is leading with your upsell instead of their goal. The moment an interview feels like a sales pitch, customers give you polite non-answers. Open with what they are trying to accomplish and let the adjacency surface on its own — the AI interviewer's job is to probe their world, not present yours. The second is interviewing only happy accounts. Your champions will tell you everything is great and you will learn nothing about the blockers stopping wider rollout; deliberately include accounts with flat or modest usage, because their growth barriers are often the most fixable. The third is treating expansion research as a one-time event. Account needs change with every reorg, new hire, and shifting priority, so a study you run once is stale within a quarter.
It helps to connect the research directly to revenue motion. The output should not sit in a slide deck — it should become a named, prioritized account list handed to customer success and sales with the customer's own words attached, plus a short list of validated adjacencies for the roadmap. Because Koji keeps every interview's structured answers aggregatable and auto-codes the open-ended ones into themes, you get that ranked, segmented output automatically rather than reconstructing it from transcripts. That is the difference between expansion research that informs a quarterly plan and expansion research that closes deals next week.
Related Resources
Related Articles
Customer Renewal Interviews: How to Reduce B2B Churn Before Renewal (2026)
A guide to renewal interviews for B2B and SaaS teams: when to run them, what to ask, and how Koji's AI interviews catch renewal risk months before the contract date — so you fix problems while there's still time to save the account.
Customer Retention Research: The Complete 2026 Playbook for Reducing Churn Before It Happens
A practitioner's guide to customer retention research — how to combine churn interviews, stay interviews, NPS follow-ups, and continuous voice-of-customer programs to reduce churn 25% or more. Includes question templates, sampling frameworks, and how AI-moderated research scales retention listening across your entire customer base.
Customer Success Interview Guide: How to Run Proactive CS Research at Scale
How customer success teams use AI-powered interviews to proactively understand account health, expansion opportunities, and churn risk — without scheduling 1:1 calls with every customer.
Feature Adoption Research: How to Interview Users Who Aren't Using Your Product
A complete guide to understanding why users ignore, avoid, or misuse features — and how to use AI-powered interviews to get honest answers at scale.
Jobs-to-Be-Done Interview Guide
Learn the JTBD interview methodology to uncover why customers switch products and what progress they're trying to make.
Structured Questions in AI Interviews
Mix quantitative data collection — scales, ratings, multiple choice, ranking — with AI-powered conversational follow-up in a single interview.