# WhyLeft — complete documentation > AI-powered exit interviews for SaaS. WhyLeft automatically emails customers after they cancel a SaaS subscription, asks one AI-written contextual follow-up question, classifies the churn reason, and surfaces the highest-impact thing to fix — with the customer quotes and revenue behind it. Generated from the same content modules that render https://whyleft.com, so this file cannot drift from the website. Last generated at build time. --- ## What WhyLeft is WhyLeft runs exit interviews for SaaS businesses, automatically. When a customer cancels, WhyLeft detects it and sends them a short, plain-text email from the founder's own address — no survey link, no branding, no tracking pixel. The customer replies in their inbox. AI reads that reply and asks one contextual follow-up question written from what they actually said, which is where a vague answer ("too expensive") becomes an actionable one ("the seat model broke when we hit five people"). Every conversation is then classified by churn reason and read against the whole history, so patterns surface on their own instead of waiting to be noticed. ### The problem it solves Churn analytics tells you a rate. A rate is the sum of many individual decisions and contains none of the reasons. The reasons exist, in recoverable form, for roughly a week after someone cancels — and that is exactly the week in which nobody has time to go and ask. WhyLeft closes that gap. ### Who it's for SaaS founders and small teams who already know their churn number and want to know what to do about it. ### What it deliberately does not do - No retention interception, save-offers, discounts or cancellation friction. - No more than two emails per interview. - No native Zapier app (outbound webhooks cover the same ground today). - No customer names or email addresses in automated AI prompts. --- ## Features ### Exit interviews URL: https://whyleft.com/features/exit-interviews Every customer who cancels gets one short, plain-text question from your address — minutes after they leave, while they still remember why. Benefits: - It doesn't look like software: No logo, no tracking pixel, no star rating, no branded template. Plain text from your address, signed with your name — the same thing you'd have typed manually. - It arrives at the only moment that works: Minutes after the cancellation, while the reason is still fresh and they're still thinking about you. A week later they've moved on and won't reply to anything. - Replying costs them nothing: There's no form to open and no page to load. They hit reply in the client they already have open, type a sentence, and they're done. #### Why not a survey? An exit survey asks someone who has just decided to stop giving you money to click a link, load a page, and fill in a form. Almost nobody does, and the people who do are systematically the angriest — which quietly skews everything you learn. An email that can be answered by replying removes every step between the question and the answer. The customer doesn't have to decide to "go and do a thing." They just answer, in their own words, at whatever length they feel like. Free text is also the point. A dropdown forces someone to pick the closest of your five guesses. An open question lets them tell you the thing you didn't think to list — which is usually the thing worth knowing. #### You control every word The default email is short and deliberately unpolished, because that's what gets replies. You can rewrite it entirely, keep your own subject line, and set the name and address it sends from. You can also run several variants at once to see which phrasing your customers actually respond to, and switch between an AI-written email that adapts to each customer's context and a fixed template you've written yourself. #### Sending is capped, on purpose A cancelled customer receives at most two messages per interview: the question and one follow-up. That limit is enforced in the send path, not by policy — the code checks before every message goes out. Every email carries an RFC 8058 one-click unsubscribe header, so a single click in their mail client stops everything permanently. No landing page, no confirmation step, no dark pattern. #### How it works 1. Cancellation detected — From your billing provider, the API, a webhook, or a CSV import. 2. Short delay — A couple of minutes, so the email doesn't collide with your provider's own cancellation receipt. 3. Email sent from your domain — Through your own verified sending domain, so it authenticates as you and lands in the inbox. 4. Reply routed back — Their reply is captured, threaded, and handed to the follow-up and tagging steps. #### In practice: The founder who kept meaning to ask You've told yourself for a year that you'll start emailing churned customers. You did it twice, got two genuinely useful answers, and then a busy month happened. WhyLeft is that intention, running whether or not this month is busy — and unlike you, it never skips the awkward ones. --- ### AI follow-ups URL: https://whyleft.com/features/ai-follow-ups "Too expensive" isn't a reason — it's an unfinished sentence. WhyLeft reads every reply and asks the one follow-up that finishes it. Benefits: - Contextual, not templated: The follow-up is written from what this specific person said. It isn't a branch in a flowchart you had to build, and it isn't the same question everyone gets. - It asks what you would have asked: Too expensive compared to what? When did it stop feeling worth it? Which report did you actually need? The questions a founder asks reading a reply at 11pm. - Still capped at two emails: One follow-up by email, and that's the limit. If a customer wants to keep talking, the conversation can move to a private chat link instead. #### What a follow-up actually changes Take a real shape of answer: "it got too expensive for what we were using it for." Filed as-is, that lands in a bucket labelled price, and the obvious conclusion is that you should charge less. Ask one more question — was there a moment it stopped feeling worth it? — and you might get: "when we hit the seat limit. Going from 3 to 5 people doubled the bill and we only needed one of them in there daily." That's not a price problem. That's a packaging problem, and the fix is a per-seat tier rather than a discount. Same customer, same initial answer, completely different roadmap decision. The difference is entirely in the second question. #### The adaptive interview Email is capped at two messages, which is the right limit for someone who has just left. But some customers want to keep talking, and for them the interview can continue on a private hosted link. There, the AI can follow the thread properly — asking several questions, going deeper where the answer is interesting, and stopping when it has what it needs rather than when it hits a message limit. #### In practice: The reason behind the reason Six customers say "missing features." Without follow-ups, that's one tag and no decision. With them, you find that four meant the same missing integration and two meant something entirely different — and you now know which of those two roadmap items is worth four customers instead of six. --- ### Insights & tagging URL: https://whyleft.com/features/insights Every conversation is classified automatically and read against everything else you've heard — so patterns surface on their own instead of waiting for you to notice them. Benefits: - Tagged the moment it arrives: Every conversation is classified by churn reason without you reading it first — pricing, missing features, competitors, onboarding friction, low usage, and more. - Read across the whole history: The analysis holds every interview in mind at once and notices that six people said the same thing in six different ways. That's the part humans are genuinely bad at. - Weighted by what it costs you: A pattern in three $19 accounts and a pattern in three $400 accounts are not the same finding. Insights carry the MRR behind them so you can rank against your actual roadmap. #### Tags are a starting point, not a verdict Classification exists so you can find things — every interview, filtered to a reason, in one click. It is not there to replace reading what people said. Every tag can be changed, and the original text is always right beside it. You are never reading a summary of a summary, and you can always check the machine's reasoning against the actual words. #### What to fix first The output isn't a chart for you to interpret. It's a specific claim — this pattern, this many customers, this much revenue — with the customer quotes it's drawn from and a confidence level attached. You can also cut the same data by segment: by plan, by tenure, by price point. Starter customers and enterprise customers usually leave for opposite reasons, and averaging them together hides both. #### In practice: The quarter you stopped guessing You have eleven things on the roadmap and room for two. Insights tells you that nine of your last twenty-four cancellations mention the same missing capability, worth roughly $1,100 MRR a quarter at the current rate — and shows you the six quotes it's reading. Now the prioritisation argument has evidence in it. --- ### Wylo Copilot URL: https://whyleft.com/features/copilot Wylo is an AI workspace that answers questions about your own customers in plain language — and shows you which interviews it read to get there. Benefits: - Questions, not query builders: "Why are annual customers leaving?" "What changed after we raised prices?" "Which accounts are most at risk?" — asked the way you'd ask a colleague. - Grounded in your interviews: Answers cite the conversations behind them. You can open the source and check whether the conclusion actually follows from what people said. - Inside your workspace boundary: Wylo goes through exactly the same authorization boundary as the rest of your dashboard. It cannot reach data in another organization, ever. #### Why a copilot rather than another dashboard Dashboards answer the questions someone anticipated when they built them. The useful question is almost always the next one — the one that occurred to you because of what the dashboard just showed. Wylo exists for that next question. It can compare segments, look for changes over time, and pull the specific conversations behind a number, without you having to find the right filter combination first. #### How it handles your customers' data The automated pipelines — tagging, follow-ups, insights — never receive a customer's name or email address. They see the reply text and non-identifying metadata only. Wylo is the one deliberate exception, and it's founder-controlled: when you explicitly ask a question about a specific customer, it can retrieve that customer's identity, inside your own authenticated workspace, through the same permission checks your dashboard uses. Anonymised data remains the default; identity is returned only when your question actually requires it. #### In practice: The board question you can now answer Someone asks why churn moved last quarter. Instead of exporting to a spreadsheet and building a pivot, you ask — and get an answer that names the pattern, quantifies it, and links the eight conversations it came from. --- ### Retention Playbooks URL: https://whyleft.com/features/playbooks Every stage from cancellation to win-back, on one canvas — what fired, what's waiting, what's off, and what each step is producing. Benefits: - A visualization, not homework: The pipeline is code that already works. The canvas is there for you to inspect and tune it — you never have to assemble a workflow before the product does anything. - Live numbers on every stage: How many cancellations entered, how many emails went out, how many replied, how many completed, how many came back. The funnel, in place, on the thing that produced it. - Tune what matters: Delays, which segments get win-back offers, whether sends need your approval — adjustable per stage without touching the rest. #### Win-back, targeted by what they actually said A generic "we miss you" email to everyone who ever cancelled is spam with a nicer subject line. The interesting version is narrower: email the customers who left over a specific thing, once you've shipped that specific thing. Because every interview is tagged and the text is kept, WhyLeft knows which customers said what. When you fix the integration gap, the people who left over integrations are a list — not a guess. You approve every send, or you let it run. Recovered revenue is attributed back to the interview it came from, so you can see whether the loop is actually paying for itself. #### In practice: Shipping the thing they left over You finally build the export people kept asking for. Instead of announcing it to everyone and hoping, you email the eleven customers who explicitly named it as their reason for leaving — with a message that references what they told you. --- ### Cancellation Capture URL: https://whyleft.com/features/cancellation-capture A short, skippable flow inside your own cancel button — so you hear the reason at the moment of the decision, from everyone, not just the people who reply to email. Benefits: - Catches the people who'd never reply: Email reaches the customers willing to write back. The in-app flow reaches everyone else — including the ones who churn silently and never open another email from you. - Always skippable: There is no save-offer, no retention wall, no confirmation maze. It asks, they can skip, they cancel either way. Anything else is a dark pattern and, in the EU, increasingly illegal. - Merges with the email interview: What they say in the flow and what they say by email become one record for the same customer, so you aren't looking at two half-pictures. #### How it goes in One script tag and a publishable key. The flow renders in an isolated shadow root, so it can't inherit or leak your CSS, and it's configured entirely from the WhyLeft dashboard — change the questions without touching your code again. You choose which domains may use your key, so a copied snippet can't be run from someone else's site. #### Deliberately not a save flow WhyLeft does not do retention interception, and this is a product decision rather than a missing feature. Making cancellation harder produces a worse business and a worse reputation, and click-to-cancel rules are tightening in exactly that direction. The flow asks a question. It never stands between a customer and the cancel button. #### In practice: The silent majority Most of the people who leave never write back to anything. The in-app flow is the only moment you'll ever have their attention — and one skippable question at that moment is worth more than ten emails afterwards. --- ### Payment Recovery URL: https://whyleft.com/features/payment-recovery A failed card isn't a customer leaving — it's a customer who's about to leave by accident. WhyLeft spots it and gives them a capped, non-annoying way to fix it. Benefits: - Detection is always on: Failed payments and at-risk revenue show up in your dashboard whether or not you turn the emails on. You can see the problem before you decide how to handle it. - A capped sequence, not a campaign: A hard limit on how many reminders anyone receives. Someone whose card expired needs a nudge, not a drip campaign. - Kept separate from real churn: Involuntary churn is a billing problem, voluntary churn is a product problem, and mixing them corrupts both numbers. WhyLeft keeps them apart. #### Why this sits next to exit interviews A meaningful share of what most SaaS businesses count as churn is nothing more than an expired card. It's the cheapest revenue you will ever recover, and it needs no product work at all. It also distorts everything else. If involuntary churn is folded into your churn rate, your churn rate is partly a measure of card expiry dates — and the exit interview data you're reading is diluted with people who never chose to leave. #### In practice: The month you found free money You look at at-risk revenue for the first time and find several hundred dollars of MRR sitting in failed payments — customers who are still using the product and have no idea anything is wrong. --- ### Integrations & API URL: https://whyleft.com/features/integrations Stripe, Polar, Lemon Squeezy and Paddle are detected automatically. Everything else works through one API call. Benefits: - Four providers, zero code: Connect the account and cancellations start flowing. Access is read-only and narrow — WhyLeft never touches card data and never modifies anything. - Any other stack, one endpoint: POST a cancellation from wherever your own cancel logic lives. Everything downstream behaves identically to a detected one. - Alerts where you already are: Slack notifications for the things worth interrupting you for: a reply landing, a high-MRR account leaving, a pattern emerging. #### The API is included There's no add-on and no enterprise tier for it. Send cancellations in, receive events out through webhooks, and export everything you've collected whenever you want it. The billing-provider integrations are a convenience, not a requirement. Plenty of businesses run entirely through the API because their cancel flow lives somewhere a webhook can't see. #### Automation without a native app WhyLeft doesn't ship a native Zapier app today. Outbound webhooks cover the same ground — point one at Zapier, n8n, Make, or your own endpoint and react to WhyLeft events wherever you like. We'd rather say that plainly than list a logo we haven't built. #### In practice: The homegrown billing stack Your subscriptions live in your own database because of an old decision nobody wants to revisit. That's fine — one POST from the function that already handles cancellation, and the rest of WhyLeft works exactly as advertised. --- ## Pricing URL: https://whyleft.com/pricing WhyLeft has ONE plan called "WhyLeft". There are no tiers and no feature gates — every subscriber, and every account inside the 14-day full-access free trial, gets every capability. - Monthly: $29/month - Annual: $278.40/year (a flat 20% discount) - Free trial: 14 days, full access, no credit card required Included in the base subscription: - 1 Organization (a completely isolated SaaS business — its own customers, interviews, insights, playbooks, templates and settings) - 3 Seats (a seat is one human, licensed at the account level, so one teammate can work across every Organization for a single seat) - Unlimited exit interviews, AI follow-ups, insights, interview history and win-back campaigns - Slack alerts, A/B email testing, CSV export, API access, custom email templates, removal of WhyLeft branding, and every future core feature Expansion is by quantity, not by tier: - Additional Organization: +$20/month each - Additional Seat: +$10/month each Both can be added or removed at any time from the billing page, with no plan switching and no re-subscribing. Cancelling is one click; access continues to the end of the paid period and data stays exportable throughout. --- ## Security and data handling URL: https://whyleft.com/security - Row-level security is enabled on every database table without exception, and every query is additionally scoped to the caller's Organization. - Automated AI processing (tagging, follow-up generation, insights, recommendations) receives reply text and non-identifying metadata only — never a customer's name or email address. This is enforced where the prompt is built. - The single deliberate exception: when the authenticated account owner explicitly asks Wylo Copilot about a specific customer, it can retrieve that customer's identity — only within their own Organization, only through the same authorization checks the dashboard uses. - Secrets (provider credentials, API keys) are encrypted at rest with AES-256-GCM. - Billing-provider access is read-only and narrowly scoped: detect cancellations and look up the cancelled customer's email. WhyLeft never reads card data and never modifies anything. - WhyLeft is incorporated in Estonia (EU) and operates under GDPR. The customer is the Data Controller for their customers' data; WhyLeft is the Processor. A DPA is available at privacy@whyleft.com. - Anthropic processes data as a sub-processor under Zero Data Retention: prompts and outputs are not retained and are never used to train models. - Sub-processors: Vercel (hosting), Supabase (database), Clerk (auth), Stripe (billing), Postmark (interview email), Resend (transactional email), Anthropic (AI, ZDR), Upstash (queue and rate limiting). - Data can be exported or permanently deleted from Settings at any time. Deletion is real deletion. Billing records are retained 7 years per Estonian tax law. --- ## API URL: https://whyleft.com/api-docs Endpoint: POST https://app.whyleft.com/api/v1/cancellations Auth: `Authorization: Bearer wl_...` (create keys in the dashboard under Integrations → API; keys are scoped to one Organization and revocable) Request fields: - customer_email (string, required) — who cancelled; the interview is sent here - customer_name (string, optional) — personalises the greeting; never sent to AI - plan_name (string, optional) — used for segmenting insights - mrr_cents (integer, optional) — monthly revenue lost, in whole cents - cancelled_at (ISO 8601, optional) — a future date schedules the interview - send_before_days (integer, optional) — for future cancellations, send N days before - idempotency_key (string, optional) — same key twice returns the same interview - send_email (boolean, optional, default true) — false records without interviewing Responses: 201 created · 200 deduplicated · 401 bad key · 403 no active subscription or pending deletion · 409 duplicate · 413 payload over 64 KB · 422 validation failure or unsubscribed recipient · 429 rate limited. Outbound webhooks push WhyLeft events to Zapier, n8n, Make or a custom endpoint. Full exports and a GDPR export endpoint are available. --- ## Frequently asked questions URL: https://whyleft.com/faq ### Getting started **How long does setup actually take?** (https://whyleft.com/faq#setup) About five minutes. You connect your billing provider, confirm the name and address your exit emails send from, and write one or two sentences describing what your product does so the AI has context. There's no code to deploy, no snippet to install, and no data migration. The next customer who cancels gets an interview. **What if I don't have many cancellations?** (https://whyleft.com/faq#no-cancellations-yet) Then WhyLeft costs you nothing to run and catches every single one, which is exactly when each individual answer is most valuable. Ten cancellations you understand completely is a better dataset than four hundred you can only count. You can also send a test cancellation from the dashboard to see the whole flow end to end before a real customer ever hits it. **How does the free trial work?** (https://whyleft.com/faq#trial) 14 days, full access to everything, no credit card required. If the trial ends before you subscribe, your data stays exactly where it is — nothing is deleted, and subscribing picks up where you left off. ### Interviews **Will my customers know it's automated?** (https://whyleft.com/faq#automated) The email is plain text, sent from your address, signed with your name, with no tracking pixel, no logo, no survey link and no unsubscribe footer beyond the legally required one-click header. It arrives minutes after they cancel, which is exactly when a founder would have written it. When they reply, the reply goes into a real conversation. Most people never think about it. **Will people actually reply?** (https://whyleft.com/faq#reply-rate) More than reply to a survey link — because it isn't one. It's one question from a human address that they can answer by hitting reply, without leaving their inbox or loading a form. We don't publish a promised reply rate, because yours will depend on your product, your relationship with your customers and how you word the question. You'll see your own rate in the dashboard from your first cancellation. **How many emails does a cancelled customer get?** (https://whyleft.com/faq#email-count) At most two: the exit interview and one follow-up. That cap is enforced in code, not by policy — the send path checks it before every message. Win-back emails are a separate, opt-in flow that only goes to customers whose stated reason you've actually addressed. **Can customers opt out?** (https://whyleft.com/faq#opt-out) Every email carries an RFC 8058 one-click unsubscribe header, so a single click in their mail client stops everything — no landing page, no confirmation step, no dark pattern. Unsubscribes are honoured immediately and permanently. **What's the hosted chat interview?** (https://whyleft.com/faq#hosted-chat) If a customer would rather have a back-and-forth than an email thread, the interview can continue on a private link. It's the same conversation, just in a chat interface, and it lets the AI ask more than one follow-up where email is capped at two messages total. ### AI **Which AI models does WhyLeft use?** (https://whyleft.com/faq#which-ai) Anthropic's Claude. Fast, cheap models handle classification and tagging; stronger models write the follow-up questions and the cross-interview analysis. Model choice is configured per workflow rather than hardcoded, so it improves as the models do. **What if the AI gets a churn reason wrong?** (https://whyleft.com/faq#ai-accuracy) You can see and change every tag. Tags are a starting point for finding patterns, not a verdict — and the original text of every conversation is always right there next to the tag, so you're never reading a summary of a summary. AI output in WhyLeft is treated as a suggestion throughout. Insights and recommendations come with the quotes they're derived from so you can check the reasoning yourself. **What is Wylo Copilot?** (https://whyleft.com/faq#copilot) An AI workspace that can answer questions about your own churn data in plain language — "why are my annual customers leaving?", "what changed since we raised prices?" — grounded in your interviews, and showing which ones it used. It only ever sees data inside your own workspace, and it goes through exactly the same authorization boundary as the rest of your dashboard. ### Pricing & billing **What does WhyLeft cost?** (https://whyleft.com/faq#price) $29/month, or $278.40/year for a flat 20% off. That's the whole price list. There is one plan and there are no feature gates: every subscriber gets unlimited interviews, unlimited AI follow-ups, unlimited insights, win-back campaigns, Slack alerts, the API, custom templates, and every core feature added in future. **What are Organizations and Seats?** (https://whyleft.com/faq#orgs-seats) An Organization is one completely isolated SaaS business — its own customers, interviews, insights, playbooks, templates and settings. If you run two products, you want two Organizations. Each additional one is $20/month. A Seat is one human. 3 are included, and a seat is licensed at the account level, so one teammate can work across every Organization you run without costing you more than one seat. Extra seats are $10/month. Both can be added or removed at any time from the billing page — no plan switching, no re-subscribing. **What happens if I cancel?** (https://whyleft.com/faq#cancel) You keep access until the end of the period you've paid for, and your data stays available for export the whole time. Cancelling is one click in the billing page — no email, no retention call, no "are you sure" maze. We'd appreciate it if you told us why. We do run our own exit interviews. ### Integrations & API **Which billing providers work with WhyLeft?** (https://whyleft.com/faq#billing-providers) Stripe, Polar, Lemon Squeezy and Paddle are detected automatically — connect the account and cancellations start flowing. Anything else works through the API, an outbound webhook from your own backend, manual entry, or a CSV import. The billing provider is a convenience, not a requirement. **I don't use any of those. Can I still use WhyLeft?** (https://whyleft.com/faq#no-stripe) Yes. POST a cancellation to the API from wherever your own cancel logic runs and everything downstream works identically. It's one endpoint, one API key and a handful of fields. **Is there an API?** (https://whyleft.com/faq#api) Yes, and it's included — no add-on, no enterprise tier. You can send cancellations in, receive events out via webhooks, and export everything you've collected. **Can I get alerts in Slack?** (https://whyleft.com/faq#slack) Yes. Connect a workspace and choose what's worth interrupting you for — a reply landing, a high-MRR account cancelling, or an emerging pattern across several interviews. ### Privacy, security & GDPR **Is WhyLeft GDPR compliant?** (https://whyleft.com/faq#gdpr) WhyLeft is incorporated in Estonia and built for GDPR from the start. You are the Data Controller for your customers' data; we are the Processor, and a DPA covering that relationship is available on request. Every email carries one-click unsubscribe, you can export or permanently delete all customer data from Settings at any time, and deletion is real deletion rather than a hidden flag. **Does my customers' personal data go to the AI?** (https://whyleft.com/faq#customer-pii) No. Automated AI processing — tagging, follow-up generation, insights, recommendations — receives the reply text and non-identifying metadata only: plan, price, dates, churn tags. Never a name, never an email address. The one exception is deliberate and founder-controlled: when you personally ask Wylo Copilot a question about a specific customer, it can retrieve that customer's identity — inside your own authenticated workspace only, through the same authorization boundary as your dashboard. **Does Anthropic train on my data?** (https://whyleft.com/faq#data-retention) No. Anthropic processes data as our sub-processor under a Data Processing Agreement with Zero Data Retention: prompts and outputs are not retained and are never used to train models. **What access does WhyLeft have to my Stripe account?** (https://whyleft.com/faq#stripe-access) Read-only, and narrowly scoped. WhyLeft uses the connection to detect cancellations and look up the cancelled customer's email address. It never reads card data, never moves money, and never modifies anything in your account. **How is my data isolated from other customers'?** (https://whyleft.com/faq#data-isolation) Row-level security is enabled on every table in the database without exception, and every query is scoped to your Organization. Secrets you give us — API keys, provider credentials — are encrypted at rest. --- ## Changelog URL: https://whyleft.com/changelog - 2026-07 [new] Wylo Copilot: Ask questions about your churn data in plain language and get answers grounded in your own interviews, with the conversations it used shown as sources. Replaces the earlier read-only analyst view. - 2026-07 [improved] Adaptive interviews and AI-written exit emails: The hosted chat interview can now follow a thread properly — asking several questions and going deeper where the answer is interesting — instead of a fixed two-turn exchange. Exit emails can also be written per-customer from your product context rather than sent from a fixed template. - 2026-07 [new] Per-interview recommendations: Individual interviews now carry their own recommendation alongside the cross-interview analysis, with the segment and the specific excerpt it was drawn from. - 2026-07 [improved] Custom sending addresses: Choose the local part of your sending address — support@, hello@, your own — once your domain is verified. Exit interviews look even more like they came from you, because they did. - 2026-07 [improved] One plan, no feature gates: Starter and Pro tiers were replaced by a single plan that includes everything. Expansion is now by Organizations and Seats, added or removed any time without switching plans. - 2026-07 [improved] Dedicated sending infrastructure: Paid accounts send on dedicated infrastructure with per-organization domain verification and delivery monitoring, so your exit interviews don't share a reputation with anyone else's. - 2026-07 [new] Retention Playbooks: The whole pipeline from cancellation to win-back, on one canvas, with live per-stage numbers and controls for delays, segments and approvals. - 2026-06 [new] Cancellation Capture: A short, always-skippable flow inside your own cancel button, so you hear the reason at the moment of the decision — including from the customers who would never reply to an email. One script tag, configured from the dashboard. - 2026-06 [new] Payment Recovery: Failed payments are detected and separated from voluntary churn, with at-risk revenue surfaced in the dashboard and a capped reminder sequence you can switch on. - 2026-06 [new] Organizations and teams: Run several isolated businesses from one account, invite teammates with roles, and switch between workspaces. Seats are licensed per human, not per workspace. - 2026-06 [improved] Polar, Lemon Squeezy and Paddle: Automatic cancellation detection beyond Stripe. Anything not on the list still works through the API, a webhook, or CSV import. - 2026-06 [new] Recovered revenue attribution: When a customer who left comes back, the resubscription is attributed to the interview that produced it — so win-back stops being a leap of faith and starts being a number. --- ## Articles ### Your churn rate is not a reason URL: https://whyleft.com/blog/churn-rate-is-not-a-reason Published: 2026-07-27 There is a specific kind of meeting that happens in every SaaS company. Churn is up forty basis points. Everyone agrees this is bad. Someone pulls up a cohort chart. Someone else suggests it might be seasonal. Forty minutes later the meeting ends, and nobody in the room has learned a single thing about why any individual human being decided to stop paying. This isn't a failure of analytics. Analytics did its job perfectly — it told you the number, segmented it, and plotted it against last quarter. The problem is that the number was never going to contain the answer, and no amount of slicing will make it. ### Rate is an outcome. Reason is a cause. A churn rate is the sum of a large number of individual decisions, each made by a person with specific reasons at a specific moment. Averaging them produces something real but useless: a measurement of the outcome, containing none of the causes. The uncomfortable part is that you can improve at measuring indefinitely without getting any closer to a decision. A dashboard that refreshes hourly instead of daily tells you the same nothing, faster. ### The reason has a shelf life Here's what makes this genuinely hard rather than merely neglected: the reason exists, in a recoverable form, for about a week. On the day someone cancels, they can tell you precisely what happened. Which feature was missing. Which competitor's demo they watched. The specific Tuesday it stopped being worth the money. Ask a month later and you get a vague, reconstructed, socially acceptable version — 'it just wasn't the right fit' — which is true and completely useless. So the window in which the most valuable data in your company can be collected is also the window in which you are least likely to think about collecting it, because someone just cancelled and you're annoyed. ### Why surveys don't fill the gap The obvious answer is an exit survey, and the obvious answer mostly doesn't work — for two reasons that compound. - Almost nobody fills them in. You're asking someone who has just decided to stop giving you money to click a link, load a page, and complete a form. The activation energy is enormous and the reward is zero. - The people who do respond aren't representative. Filling in an exit survey requires a strong feeling. Mild disengagement — the single most common reason people leave — produces no strong feeling at all, so it's systematically absent from your data. - Fixed options force people into your existing guesses. A dropdown can only return an answer you already thought of, which means it structurally cannot tell you the thing you don't know. You end up with a small, biased, pre-categorised sample and a quiet, false sense that you've measured something. ### The second question Even when you do get a free-text answer, the first one is almost never the real one. Not because people lie, but because first answers are short by default. > It got too expensive for what we were using it for. Filed as-is, that's a price complaint, and the obvious response is a discount. Ask one more question — was there a moment it stopped feeling worth it? — and you might get this: > When we hit the seat limit. Going from 3 to 5 people doubled the bill and we only needed one of them in there daily. That is not a price problem. It's a packaging problem, and the fix is a viewer seat, not a discount. Same customer, same opening answer, completely different roadmap decision — and the entire difference is one follow-up question that most companies never ask because there's nobody available to ask it. ### What to actually do None of this requires a tool. It requires that three things happen reliably, and reliability is the hard part: - Every cancelled customer gets asked, within hours, in a way that costs them nothing to answer. Not a link — a question they can reply to. - Every answer gets one follow-up, written from what they actually said rather than from a script. - Every conversation ends up somewhere the whole set can be read at once, because six people saying the same thing across eight months is invisible in an inbox. Do that by hand for a quarter and you will learn more about your product than your dashboard has told you all year. Most founders manage it for about two weeks — which is precisely the gap WhyLeft exists to fill. --- ## All pages - https://whyleft.com — Home: WhyLeft runs AI-powered exit interviews for SaaS: it talks to customers after they cancel and turns the answers into things you can fix. - https://whyleft.com/features — Features: Everything WhyLeft does, from the first exit email to the win-back. - https://whyleft.com/demo — Demo: The real WhyLeft dashboard on a sample workspace — the same product paying customers use, filled with fictional data. No sign-up. - https://whyleft.com/how-it-works — How it works: The full path a cancellation takes through WhyLeft — detection, interview, follow-up, root cause, recommendation, win-back. - https://whyleft.com/pricing — Pricing: One plan, every feature included, 14-day free trial. Expansion is by Organizations and Seats, not by tiers or feature gates. - https://whyleft.com/faq — FAQ: Answers on AI, security, pricing, data, privacy, interviews, integrations, billing, GDPR, cancellation tracking and the API. - https://whyleft.com/security — Security: How WhyLeft protects your data and your customers': row-level isolation, encrypted secrets, read-only billing access, and no customer PII in automated AI prompts. - https://whyleft.com/api-docs — API: Send cancellations from any billing stack: endpoints, authentication, payloads and webhooks for the public WhyLeft API. - https://whyleft.com/blog — Blog: Writing on churn, retention, and building WhyLeft in public. - https://whyleft.com/changelog — Changelog: What shipped, when. - https://whyleft.com/reviews — Reviews: What people say about WhyLeft. - https://whyleft.com/about — About: Who builds WhyLeft, and why it exists. - https://whyleft.com/contact — Contact: How to reach support, privacy, and legal. - https://whyleft.com/privacy — Privacy: The WhyLeft privacy policy. - https://whyleft.com/terms — Terms: The WhyLeft terms of service. - https://whyleft.com/docs — Documentation index: Machine-readable index of every WhyLeft page, feature and document, including llms.txt and the markdown twins. - https://whyleft.com/features/exit-interviews — Exit interviews: A plain-text email from your address, minutes after a customer cancels. No survey link, no form — a question they can just reply to. - https://whyleft.com/features/ai-follow-ups — AI follow-ups: Every reply gets one contextual follow-up question, written from what they actually said. Continue by email or in a hosted chat. - https://whyleft.com/features/insights — Insights & tagging: Every conversation is tagged by churn reason automatically, then read across your whole history to surface the highest-impact thing to fix. - https://whyleft.com/features/copilot — Wylo Copilot: Ask your churn data questions in plain language and get answers grounded in your own interviews, with the sources it used. - https://whyleft.com/features/playbooks — Retention Playbooks: The pipeline from cancellation to win-back, visible as a canvas you can inspect and tune — not a builder you have to assemble. - https://whyleft.com/features/cancellation-capture — Cancellation Capture: A short, skippable flow inside your own cancel button — so you hear the reason before they're gone, not after. - https://whyleft.com/features/payment-recovery — Payment Recovery: Failed payments are churn you didn't choose. WhyLeft spots them and runs a capped reminder sequence to recover the revenue. - https://whyleft.com/features/integrations — Integrations & API: Stripe, Polar, Lemon Squeezy and Paddle detected automatically. Anything else works through the API, webhooks, or CSV import. ## Contact - Support: support@whyleft.com - Privacy and data requests: privacy@whyleft.com - Legal: legal@whyleft.com - Application: https://app.whyleft.com