{"site":{"name":"Koji","description":"AI-native customer research platform that helps teams conduct, analyze, and synthesize customer interviews at scale.","url":"https://www.koji.so","contentTypes":["blog","documentation"],"lastUpdated":"2026-11-11T19:57:20.448Z"},"content":[{"type":"blog","id":"86b39fe7-bc0b-4f59-b8a2-759ce88a1459","slug":"saas-packaging-change-customer-interviews","title":"How to Change SaaS Packaging Without Backlash: A Pricing Team Playbook for Customer Interviews Before You Move Features Between Tiers","url":"https://www.koji.so/blog/saas-packaging-change-customer-interviews","summary":"A playbook for SaaS pricing and monetization teams on researching a packaging change before launch. Six weeks out, interview three groups (20 to 30 active users of the feature on the tier it leaves, 10 to 15 customers on the target tier, 10 to 15 recent buyers of the lower tier), label each feature a leader, filler or killer (Ramanujam and Tacke, Monetizing Innovation), run a price test on the target tier with 30 or more affected customers, then decide with three questions: is it a reason they bought, do they know what they would do without it, and would most buy the higher tier at its price. Give at least 60 days of notice and interview again 30 days after launch. Context: three in four software companies changed pricing or packaging in the last year and hybrid pricing reached 37% (Growth Unhinged, 2026); average price realization is 43% (Simon-Kucher, 2025). Includes an eight-question interview guide and the setup in Koji.","content":"To change SaaS packaging without a backlash, interview the customers who use the feature you plan to move before the new plans go live. Six weeks out, talk to 20 to 30 of them, plus the people already on the tier it is moving to, and learn three things: whether the feature is a reason they bought, what they would do without it, and what upgrade price they would still accept. AI-moderated interviews let a pricing team hear all of them in the same week, with a follow-up on every answer, so the decision rests on what customers say rather than on what the loudest account manager remembers. By the end of this playbook you will have a six-week method, an interview guide to copy, a rule for deciding whether a feature moves, and the exact setup in Koji.\n\n## Why packaging changes need customer interviews now\n\nPackaging is moving faster than it ever has. In Kyle Poyar's [2026 State of B2B Monetization survey](https://www.growthunhinged.com/p/the-state-of-b2b-monetization-in-2026) of about 230 software and AI companies, run in April and May 2026, three in four had changed their pricing or packaging in the last year. Hybrid pricing (a seat price with usage or AI credits on top) became the most common model, at 37%, up from 25% a year earlier, and 29% now let customers choose between several pricing models, up from 21%.\n\n![Bar chart comparing software companies a year earlier and in 2026: hybrid pricing rose from 25% to 37%, and offering several pricing models from 21% to 29%.](/visuals/software-companies-hybrid-pricing-2025-2026/software-companies-hybrid-pricing-2025-2026.png \"Most of these moves change what sits in which plan · Source: Kyle Poyar, Growth Unhinged, 2026 State of B2B Monetization survey, 2026\")\n\nAI is the reason most of these changes happen. [Metronome's summary of the same report](https://metronome.com/blog/2026-trends-from-cataloging-50-ai-pricing-models) says 29% of software companies have adopted an AI credit or token model and another 33% plan to within 6 to 12 months. Each of those moves is a packaging decision: which features stay in the seat price, which go into credits, and which move up a tier.\n\nBuyers have noticed. In June 2026 Gartner published [Runaway AI Costs](https://www.gartner.com/en/documents/8034533), which tells procurement teams to stop treating embedded AI like a routine renewal and to unbundle it. Its abstract says: \"Software vendors aggressively bundle and embed AI into SaaS, inflating prices and forcing buyers to pay premiums for features that often result in shelfware.\" Your next repackaging will land on a desk that has been told to push back.\n\nAnd the money you plan for often doesn't arrive. Simon-Kucher's [Global Pricing Study 2025](https://www.simon-kucher.com/en/insights/global-pricing-study-2025), a survey of more than 2,200 business leaders in 28 countries, puts the average price realization rate at 43%, down 5 points, and names customer resistance and competitive pressure as the main barriers. A tier move is a price increase for everyone who needs the feature that moved. Guess wrong about who needs it and you pay twice: the upgrade revenue doesn't come, and the accounts that feel cheated start looking around. Skip the interviews and you find out which feature was a reason to buy from the support queue.\n\n## What good looks like\n\nTeams that repackage well treat it as research with a deadline, not a spreadsheet exercise. These are the habits that separate them:\n\n![A small team around a table with laptops, working through plan options on a whiteboard before a packaging change](https://images.unsplash.com/photo-1681949215173-fe0d15c790c1?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=M3w5NDUzMDN8MHwxfHNlYXJjaHw5fHx0ZWFtJTIwbWVldGluZyUyMHdoaXRlYm9hcmQlMjBsYXB0b3BzfGVufDF8MHx8fDE3OTQzODM0MTB8MA&ixlib=rb-4.1.0&q=80&w=1080)\n\n_Photo by [Sable Flow](https://unsplash.com/@sableflow?utm_source=koji&utm_medium=referral) on [Unsplash](https://unsplash.com/?utm_source=koji&utm_medium=referral)_\n\n- **They name the move precisely.** \"Move SSO from Pro to Business for new and existing customers on 1 March\" can be tested. \"Rework our tiers\" cannot.\n- **They interview by group, not by account size.** The people who use the feature on the lower tier, the people already on the higher tier and recent buyers of the lower tier tell you different things. Mix them and the answers average into nothing.\n- **They sort features into leaders, fillers and killers.** The framework comes from Madhavan Ramanujam and Georg Tacke's _Monetizing Innovation_ (Wiley, 2016). A leader is a reason to buy the tier. A filler is nice to have. A killer is something people refuse to pay for. Moving a leader out of a tier breaks the tier. Moving a filler up is usually fine. A killer belongs in an add-on.\n- **They ask about behaviour before price.** What someone did last week with the feature is evidence. What they say they would pay is a stated preference and needs a second check.\n- **They write the notice from the interviews.** The words customers use for the feature, and the reasons they gave, go straight into the announcement.\n\n### Common mistakes\n\n1. Interviewing only your happiest customers, who will accept anything.\n2. Asking \"would you pay more for this?\" with no price and no alternative.\n3. Leading with the change, so every interview turns into a negotiation.\n4. Treating a stated price as a forecast. People overstate what they would buy.\n5. Deciding before the interviews finish because the pricing page is already designed.\n6. Skipping the follow-up interviews 30 days after launch, when the real reactions arrive.\n\n## The Tier Move Test: six weeks back from launch\n\nThis is the method we would give any pricing team planning to move a feature between tiers or into AI credits. Count back six weeks from the date the new plans go live.\n\n1. **Week 1: write the move as a hypothesis.** One line: which feature, from which tier, to which tier, for whom, from what date, and the revenue you expect. Pull everyone who used the feature in the last 90 days from your product data.\n2. **Week 1: pick three groups.** 20 to 30 active users of the feature on the tier it is leaving. 10 to 15 customers already on the tier it is moving to, to learn why they upgraded. 10 to 15 customers who bought the lower tier in the last six months, to learn what made them choose it.\n3. **Weeks 2 and 3: run the interviews.** One study per group, with the same core questions, so you can compare them side by side.\n4. **Week 3: classify every feature in the tier.** Leader, filler or killer, for each group, using what people said and ranked.\n5. **Week 4: price the target tier.** Run a price test on the tier the feature is moving to, with 30 or more people from the affected group, and read the reasons behind each answer.\n6. **Week 5: decide and write the notice.** Use the decision path below. Give customers at least 60 days' notice, in their own words for the feature.\n7. **Launch plus 30 days: interview again.** Talk to affected accounts who upgraded, stayed or left, and fix the next move before you make it.\n\n![Process showing the Tier Move Test in seven steps, from writing the move as a hypothesis to interviewing affected accounts 30 days after launch.](/visuals/tier-move-test-packaging-change-steps/tier-move-test-packaging-change-steps.png \"Six weeks back from the day the new plans go live\")\n\n## Should this feature move? A decision rule\n\nAfter week 3 you will have dozens of interviews. Use this path to turn them into a decision for each feature on your list.\n\n1. **Is it a reason they bought the lower tier?** If most of the active users in group one name it unprompted, or rank it first or second, it is a leader. Keep it where it is and look for a different fence.\n2. **Do they know what they would do without it?** If they can't say, probe the workaround: what they used before it existed and what that cost them. That tells you how much it is really worth.\n3. **Would most of them buy the higher tier at its planned price?** If the price test shows most of the affected group would not buy the higher tier at its planned price, move the feature for new customers only, keep current customers on their plan, and test again in six months.\n4. **If all three pass, move it.** Give 60 days' notice, say plainly what they keep, and name a person to contact.\n\n![Flowchart for deciding whether a feature should move to a higher tier: three yes or no questions, one side step, ending in a follow-up 30 days after launch.](/visuals/should-this-feature-move-tiers-flowchart/should-this-feature-move-tiers-flowchart.png \"Run every feature on your list through the same three questions\")\n\n## How to set it up in Koji, step by step\n\n1. **Start a study.** Type your one-line brief (below) into the \"What do you want to research?\" box on the home screen, or open the template gallery and start from **Willingness-to-Pay Research**, which already asks what people spend today and what drove the price they agreed to pay.\n2. **Shape the research brief.** The research assistant drafts the goal, the questions and the follow-up depth. Upload your pricing page or plan comparison as a study context file so the interviewer knows what each tier contains.\n3. **Add guardrails.** In the brief, list what the interviewer must not do: quote the new price before the price test, say the change is decided, or discuss other customers.\n\n![Koji guardrails settings listing topics the AI interviewer should avoid, such as quoting prices and naming other customers](/email/shots/shot-guardrails.jpg \"Guardrails keep the plan you haven't announced out of the conversation\")\n\n4. **Add structured questions where you need a number.** A ranking question for the features in the tier (turn on option shuffling so the first item isn't favoured), a single choice question for what they would do without the feature, a 1 to 10 scale for fairness, and a **Price Test (Gabor-Granger)** question for the target tier with a ladder of 5 to 7 prices. Each one gets a follow-up asking why.\n5. **Choose voice or text.** Text suits busy admins who answer between meetings. Voice gets longer, more candid stories. You can turn on both and let people pick.\n6. **Invite your customers.** Import the list from your product data as a CSV in Participants, so each person gets a private invite link, and send the recruitment email with reminders from Koji. Add screening questions so only people who use the feature reach the interview. Duplicate the study for each of the three groups.\n7. **Read the report.** Each group's report gives you the themes with the share of people behind each one and the quotes that support it, a chart for every structured question, and the share who would buy at each price with the reasons. Download it as a PowerPoint deck or PDF for the pricing committee.\n\nYou can build and preview the whole study on a free account, which includes 10 interviews, and the price test works on every plan. Sending recruitment emails and paying incentives needs a paid plan; see [pricing](/pricing). For the price question itself, [the price test setup guide](/docs/price-test-setup-guide) explains every setting.\n\n## Copy this interview guide\n\n**Paste this into Koji's \"What do you want to research?\" box:** Find out whether customers on [Pro] treat [feature] as a reason they bought, what they would do if it moved to [Business], and what price for [Business] they would still accept, so we can decide the new packaging before [date].\n\n**Opening line:** \"Thanks for making time. We're reviewing how our plans fit the way people actually work, and your experience with [product] will shape it. There are no wrong answers.\"\n\n1. **Open:** \"Walk me through the last time you used [feature]. What were you doing, and what happened next?\" Probe: \"How often does that come up?\"\n2. **Open:** \"When you chose [Pro], what made it the right plan for you?\" Probe: \"What almost made you choose a different one?\"\n3. **Ranking:** \"Put these parts of [Pro] in order of how much you would miss them.\" Probe: \"Why is the top one first?\"\n4. **Open:** \"Is there anything in your plan you pay for and rarely use?\" Probe: \"Would you rather it were optional?\"\n5. **Single choice:** \"If [feature] were only in [Business], what would you most likely do?\" Options: upgrade to keep it, use a workaround, look at other tools, it wouldn't affect me, not sure. Probe: \"What would the workaround be?\"\n6. **Price test:** \"[Business], including [feature]: how likely would you be to buy it at this price?\" Probe: \"What makes that price feel right or wrong?\"\n7. **1 to 10 scale:** \"If this change came with 60 days' notice, how fair would it feel?\" Probe: \"What would make it feel fairer?\"\n8. **Open:** \"If your plan changed, what would you want us to tell you, and who else on your team would need to hear it?\"\n\n**Closing:** \"What should we have asked about your plan that we didn't?\"\n\n![Question bank for packaging change interviews in three sections: how they use the feature today, what the plan is worth to them, and how the change would land.](/visuals/packaging-change-interview-questions/packaging-change-interview-questions.png \"Eight questions to reuse, in order\")\n\n## What you get\n\nAt the end of week 3 you have three short reports, one per group, and a leader, filler or killer label for each feature in the tier, backed by quotes. At the end of week 4 you have the share of affected customers who would buy the higher tier at each price, and the price that earns the most from them. The report labels this as directional, which is right: it tells you where demand falls off, not a forecast.\n\nThe next day, take the deck to the pricing committee to show which moves the evidence supports. Give sales and customer success the phrases customers used, so the announcement sounds like someone listened. Hand finance an upgrade rate grounded in what the affected group said, not the one in the original model.\n\n## The same playbook elsewhere\n\nCustomer success teams run a version of this around a list price increase: one wave before the letter and one after. Our [price increase rollout playbook](/blog/price-increase-rollout-interviews-customer-success-playbook) covers it. Product managers pricing a new AI feature ask many of the same questions before it has a tier at all; see [pricing an AI feature](/blog/ai-feature-pricing-research-product-managers). Deal teams test whether a target company can push through the price rises in its model with the [pricing power due diligence playbook](/blog/pricing-power-due-diligence-playbook). Borrow the deal team's habit of writing the claim down before the interviews start, and the CS habit of a second wave after launch.\n\nIf you are designing tiers from scratch rather than moving one feature, start with [good-better-best tier research](/docs/good-better-best-pricing-research). If you need to rank 20 or more features by value, a [MaxDiff study](/docs/maxdiff-analysis-guide) is the better first step, and interviews explain the top of the list.\n\n## Why Koji for packaging change interviews\n\n|                             | Koji AI interviews                 | Survey with MaxDiff   | Conjoint study           | Calls by your team             |\n| --------------------------- | ---------------------------------- | --------------------- | ------------------------ | ------------------------------ |\n| Hears why                   | Yes, a follow-up on every answer   | No, ranks only        | No, models trade-offs    | Yes, when there is time        |\n| Reach                       | Many people at once, voice or text | Many people           | Many people              | One call at a time             |\n| Same core questions for all | Yes, plus follow-ups               | Yes                   | Yes                      | Varies by caller               |\n| Price people still accept   | Built-in price test with reasons   | Needs extra questions | Yes, modelled            | Hard to ask the same way twice |\n| Analysis effort             | Themes and quotes drafted for you  | Charts, no reasons    | Needs a specialist       | Notes to code by hand          |\n| Best for                    | Why and how much, per group        | Ranking many features | Precise trade-off models | Your largest accounts          |\n\n![Comparison table of four ways to research a packaging change: AI interviews, a MaxDiff survey, a conjoint study and calls by your own team.](/visuals/packaging-change-research-methods-compared/packaging-change-research-methods-compared.png \"Four ways to research a tier move, compared\")\n\nThe reasons it fits this job:\n\n- **Numbers and reasons in one interview.** Ranking, choice, scale and price test questions sit next to open questions, and each structured answer gets a follow-up asking why.\n- **A price test built for the target tier.** You enter 3 to 10 prices, Koji starts each person at a random price in the middle of the ladder, finds the highest price they would still buy at, then asks why. The report shows demand and implied revenue at each price.\n- **Your own customers, reached properly.** Import your list, send private invite links and reminders, and screen out people who don't use the feature.\n- **Control over what the interviewer says.** Guardrails keep the unannounced plan, the new price and other customers out of the conversation.\n- **A report the committee can read.** Every finding opens to the quotes behind it, and the whole report downloads as editable slides or a PDF.\n\n## Questions you might have about AI interviews\n\n**Will customers read the interview as a warning that we're taking features away?** Some will wonder, so don't hide what you're doing. Say you are reviewing how plans fit the way people work, which is true, and use guardrails so the interviewer doesn't present the change as decided. Your customers' IT and finance teams are already braced for price changes: Zylo's [2026 SaaS Management Index](https://zylo.com/news/2026-saas-management-index) found that 79% of IT leaders saw price increases at renewal in the past year, 78% hit unexpected charges tied to AI features or usage-based pricing, and 61% had to cut projects because of unplanned SaaS costs. Being asked first is more likely to build trust than a surprise letter.\n\n![Bar chart of IT leaders hit by SaaS cost increases in the past year: 79% saw renewal price rises, 78% unexpected AI or usage charges, 61% cut projects.](/visuals/it-leaders-saas-renewal-price-increases-2026/it-leaders-saas-renewal-price-increases-2026.png \"Your customers' IT and finance teams are already braced for your letter · Source: Zylo, 2026 SaaS Management Index, 2026\")\n\n**Isn't stated willingness to pay unreliable?** Yes, on its own. That is why the method asks about behaviour first, why the price test is labelled directional, and why the decision rule only uses price as the third check. Pair it with your own data: upgrade rates on similar moves, and how often the feature is used.\n\n**Is 30 interviews enough for a pricing decision?** For a price test, about 30 per group is where the result starts to read clearly. For finding out whether a feature is a leader, you will usually see the pattern well before that. If a group has fewer than 10 accounts, call them yourself.\n\n**How is the data handled?** You can turn on a per-study setting that removes names, email addresses, phone numbers and employers from each transcript when the interview ends. Each interview gets a quality score, and only interviews scoring 3 or more out of 5 enter the report. More on [how Koji's AI interviewer works](/docs/how-ai-interviewers-work) and on [data privacy and security](/docs/ai-interview-data-privacy-security).\n\n## When Koji may not be the right fit\n\nIf you need a precise share-of-preference model across many features and price points, for example to simulate a dozen tier designs, a conjoint study run by a specialist is the better tool, and interviews are best used before it to choose what goes in. If only a handful of enterprise accounts use the feature and their contracts name it, this is a conversation for their account team and your legal team, not a study.\n\n## Metrics to track\n\n- Upgrade rate among affected accounts at 30, 60 and 90 days after launch\n- Downgrade and churn rate of affected accounts against everyone else\n- Support tickets and sales objections that mention the change\n- Price realization on the target tier against plan\n- Net revenue retention of the affected group over the next two quarters\n- Share of interviewees who named a workaround, checked against what they actually did\n\n## FAQ\n\n**How many customers should I interview before moving a feature to a higher tier?** 20 to 30 active users of the feature on the tier it is leaving, 10 to 15 on the target tier and 10 to 15 recent buyers of the lower tier, with at least 30 from the affected group in the price test.\n\n**How long does a packaging change study take?** Six weeks back from launch: two weeks of interviews, then a week each to classify features, run the price test and decide, plus a follow-up wave 30 days after launch.\n\n**Should I grandfather existing customers when I move a feature?** If the price test shows most affected customers would not buy the higher tier at its planned price, yes: move the feature for new customers only and test again in six months. If they would, give at least 60 days' notice instead.\n\n**What is the difference between leaders, fillers and killers?** It is a packaging framework from _Monetizing Innovation_ by Madhavan Ramanujam and Georg Tacke. Leaders are reasons to buy a tier, fillers are nice to have, and killers are features people refuse to pay for. Keep leaders, move fillers with care, and put killers in add-ons.\n\n**Can I use this method when moving features into AI credits?** Yes. Treat the credit pack as the target tier: ask how people use the feature today, what they would do if it used credits, and run the price test on the pack. Add a question on whether they could predict their monthly bill.\n\n## Ready to start?\n\nOpen the template gallery and start from **Willingness-to-Pay Research**, or paste the one-line brief above into Koji and let the research assistant build the study. See every playbook for pricing teams in [the use-case library](/use-cases/pricing-monetization), and how product teams use Koji on [Koji for product managers](/for/product-managers).\n","category":"Use Cases","lastModified":"2026-11-11T07:50:10.766629+00:00","metaTitle":"How to Change SaaS Packaging Without Backlash","metaDescription":"A pricing team playbook for interviewing customers before moving features between tiers or into AI credits: a six-week method, questions, a decision rule.","keywords":["saas packaging change","change saas packaging","tier migration","pricing and packaging research","feature gating","customer interviews pricing","ai credits packaging"],"aiSummary":"A playbook for SaaS pricing and monetization teams on researching a packaging change before launch. Six weeks out, interview three groups (20 to 30 active users of the feature on the tier it leaves, 10 to 15 customers on the target tier, 10 to 15 recent buyers of the lower tier), label each feature a leader, filler or killer (Ramanujam and Tacke, Monetizing Innovation), run a price test on the target tier with 30 or more affected customers, then decide with three questions: is it a reason they bought, do they know what they would do without it, and would most buy the higher tier at its price. Give at least 60 days of notice and interview again 30 days after launch. Context: three in four software companies changed pricing or packaging in the last year and hybrid pricing reached 37% (Growth Unhinged, 2026); average price realization is 43% (Simon-Kucher, 2025). Includes an eight-question interview guide and the setup in Koji.","aiKeywords":["saas packaging change","tier move test","leaders fillers killers","moving features between tiers","ai credits packaging","grandfathering customers","price test","pricing team customer interviews"],"aiContentType":"guide","faqItems":[{"answer":"20 to 30 active users of the feature on the tier it is leaving, 10 to 15 on the target tier and 10 to 15 recent buyers of the lower tier, with at least 30 from the affected group in the price test.","question":"How many customers should I interview before moving a feature to a higher tier?"},{"answer":"Six weeks back from launch: two weeks of interviews, then a week each to classify features, run the price test and decide, plus a follow-up wave 30 days after launch.","question":"How long does a packaging change study take?"},{"answer":"If the price test shows most affected customers would not buy the higher tier at its planned price, yes: move the feature for new customers only and test again in six months. If they would, give at least 60 days' notice instead.","question":"Should I grandfather existing customers when I move a feature?"},{"answer":"It is a packaging framework from Monetizing Innovation by Madhavan Ramanujam and Georg Tacke. Leaders are reasons to buy a tier, fillers are nice to have, and killers are features people refuse to pay for. Keep leaders, move fillers with care, and put killers in add-ons.","question":"What is the difference between leaders, fillers and killers?"},{"answer":"Yes. Treat the credit pack as the target tier: ask how people use the feature today, what they would do if it used credits, and run the price test on the pack. Add a question on whether they could predict their monthly bill.","question":"Can I use this method when moving features into AI credits?"}],"relatedTopics":["pricing and packaging","SaaS pricing","tier migration","AI credits","customer interviews","willingness to pay"]}],"pagination":{"total":1,"returned":1,"offset":0}}