The Product Succeeds When the Plan Is Finished, and That Is the Churn Problem: A Boldin PlannerPlus Retention Teardown

Executive Summary

Industry Trends

Boldin is not selling entertainment or convenience. It is selling clarity about the one financial decision that has no do-over, to people who are anxious about it and often close enough to the date that the anxiety is concrete. That changes which parts of the standard retention playbook are usable and which are actively harmful. Three structural shifts define the environment here.

First, the advice gap is being filled by software, and the software is being commoditized by AI. A comprehensive plan from a CFP runs roughly $2,000 to $5,000; Boldin sells the software for $144 a year and a written plan from its own CFPs for a $3,200 flat fee.112 That gap is exactly where the category lives, and it is filling up fast: ProjectionLab, MaxiFi and Pralana all sit within $20 of Boldin's price, and general-purpose AI assistants now answer the surface-level version of most retirement questions for free. The durable moat is not the calculation, it is the maintained plan: the accumulated inputs, the linked accounts, the scenario history, and the record of what you decided and why. Every retention mechanic worth building here compounds that asset.

Second, the audience is entering the densest decision window of its financial life, on a calendar that is public knowledge. Roth conversion opportunities cluster in the low-income years between retirement and Required Minimum Distributions. IRMAA brackets look back two years, so a conversion in one year sets a Medicare premium two years later. Age 59½, 62, 65, 67 and 73 are all hard gates. Open enrollment runs in the fall, tax filing in the spring, and ACA subsidy cliffs matter every single year for early retirees. A retirement planning product has something almost no consumer subscription has: a legitimate, non-manufactured reason to reach out on a known schedule, personalized to the user's own age and plan. Very little of that appears to be built.

Third, trust is the constraint, and Boldin's business model is its biggest asset inside it. This is a category where the incumbent alternative, a commissioned advisor, has a conflict of interest baked into the relationship. Boldin's founder makes the point directly, that the product is "100% aligned with the end user, both in terms of our business model and our product," and the AI launch repeated it: answers "grounded in their actual financial plan, without sales pressure or hidden incentives."45 That alignment is the reason a subscriber will hand over their entire net worth, and it is also a hard boundary on growth tactics. Nothing that feels like a nudge toward a product Boldin profits from is available here, which rules out most of the aggressive lifecycle playbook and makes the honest mechanics, showing people their own numbers at the right moment, the only ones on the table.

The Company

Boldin's own framing for why it exists, from the rebrand: the company had outgrown "NewRetirement" because the plan people need starts long before retirement, and the aim was to become "the most comprehensive, affordable, and widely accessible financial planning platform."4 That is a strategically sound repositioning, and it also widened the aperture on who the subscriber is. A 42-year-old modeling a home purchase and a 66-year-old sequencing withdrawals are on completely different decision calendars, which matters more for retention than for acquisition. The retention framework has to segment on that, not on tenure alone.

Their Plans

Working reference assembled from boldin.com's pricing, classroom, coaching and community pages, captured August 2026, since pricing pages change over time and the findings below cite specific numbers from them.1 Prices and features subject to change.

The portfolio, side by side

The July 2025 price change, and why it matters for retention

Competitor Analysis

The competitive set here is unusually crowded at almost exactly the same price, which tells you price is not the battleground. Retention is.

1. The product's success condition is its churn trigger, and nothing schedules the next decision

Boldin's core job is "help me figure out my retirement." That job completes. The user enters their accounts, runs their scenarios, learns their number, and arrives at the feeling the product promised: confidence. Then the reason to open it again is gone. This is not a hypothesis I invented, it is what subscribers say in public when nobody from the company is asking. One long-time forum user described paying for a single year to model his situation and deliberately planning to lapse until his circumstances changed: he would "not renew for a couple of years until I get off of ACA with CSRs then sign back up once on Medicare." Another stopped after generating five scenarios, because his "financial die is already cast."11

Read those two quotes as a PM and they are not the same churn. The first person is not lost, he is paused, and he has told you the exact re-entry date. The second is genuinely done with the job as currently defined. Blended churn treats them identically. That is the first thing I would want to split, because the fix for one is a dormancy and win-back architecture with a known trigger date, and the fix for the other is a product that has a second job.

The second job is the planning habit itself. And retirement planning has an unusually rich, non-manufactured reason to return, because the decisions arrive on a calendar the product already knows. Boldin holds each subscriber's age, filing status, account types, retirement date, and state. From that it can derive, without asking, that a subscriber turns 63 in March and therefore that this year's Roth conversion sets an IRMAA bracket two years out; that another is 72 and has one year left before RMDs start; that a third is retiring in a low-income year that will never repeat. Those are not "we miss you" emails. They are "here is the decision in front of you this quarter, here is what your own plan says about it."

The raw material is there. Boldin already ships personalized alerts, capped at 5 on Basic and unlimited on PlannerPlus, runs live classes on exactly these topics most weekdays, and now has an AI assistant that beta users touched 8 times a day.165 What appears to be missing is the connective tissue: a per-subscriber decision calendar that sequences those surfaces into a year, so the plan is revisited on purpose rather than when someone remembers.

2. Annual-only billing puts the entire year's retention decision in one moment, and the best asset available at that moment is invisible

PlannerPlus is $144 billed annually after a 14-day trial. There is no monthly option advertised anywhere on the pricing page.1 That is a defensible choice, and probably the right one for this category: annual billing removes twelve chances to cancel and matches a product whose value accrues over a year. But it has two consequences a retention PM has to design around.

First, the renewal is a single high-stakes event, and it lands at the low point of perceived value. By month twelve, the plan is built, the questions are answered, and the last memorable interaction may have been months ago. The subscriber is deciding whether to spend $144 based on how they feel that week, not on the year. Everything in Findings 1, 4 and 5 exists to make that week feel different.

Second, a meaningful share of annual churn is not a decision at all. Annual renewals fail on expired cards at a much higher rate than monthly ones, for the obvious reason that a year is long enough for a card to be reissued, and the recovery window is a single charge attempt rather than a monthly retry. Involuntary churn is the least glamorous retention work available and usually the highest ROI per engineering hour, because the subscriber already wanted to stay. I would want the voluntary and involuntary split before anything else, because if involuntary is a large slice, dunning and card-updater work outranks every roadmap item below.

Third, and this is the one I would test first: Boldin already granted every existing subscriber a permanent price lock and appears never to mention it again. In July 2025 the price went from $120 to $144 for new subscribers only, with existing members held at their current cost "for as long as your subscription stays active."2 That is an honest, real, growing switching cost: cancelling forfeits it forever, and a returning subscriber pays the new price. Most members almost certainly have no idea. Surfacing "you are locked at $120, and returning later would cost $144" at renewal and in the cancellation flow is not a dark pattern, it is a material fact the member is entitled to know, and it is free.

3. The 14-day trial measures whether someone can finish data entry, not whether the plan changed a decision

Boldin's trial is 14 days.1 Getting foundational information into the system takes 30 to 45 minutes by a sympathetic reviewer's own estimate, against a product that accepts 250+ inputs, and the most consistent criticism in every serious review is the learning curve: "the tool is not for novices."9 So the trial period is largely consumed by setup. The activation event that actually predicts renewal is almost certainly not "completed profile." It is the first moment the plan tells the subscriber something they did not already believe: a Roth conversion window they did not know they had, a claiming age that adds years of solvency, a spending number that is safe.

That distinction matters because it changes what the trial is for and what onboarding optimizes. If activation is "profile complete," onboarding optimizes for form completion and the product will keep adding inputs. If activation is "first decision-changing insight," onboarding optimizes for the shortest honest path to a surprising number, and the remaining 200 inputs become progressive refinement the subscriber does over months, which is itself a return reason.

There is a second, subtler problem. A trial is a bet that the value arrives inside the window. For a 62-year-old deciding when to claim Social Security, the window that matters is not 14 days, it is the weeks it takes to gather statements, talk to a spouse, and change their mind twice. Retirement decisions are made jointly and slowly. I would want to know how many trials end with a profile that is more than half empty, and whether the households that convert are the ones who added a second person to the plan.

4. The community Boldin is proudest of lives on Facebook and Reddit, where it cannot become a return moment

Boldin leads with its community as one of its strongest assets: "Boldin has an active community forum with users who write to us unprompted to say the platform changed their retirement." That is a genuine asset and rarer than it sounds. But Boldin's own community page points to a private Facebook group and a subreddit.8 Both are off-platform.

For a retention role, that placement has three costs. The engagement cannot be instrumented against the subscription, so "community participation" can never enter a cohort analysis or a churn model. The community cannot be a return moment, because a notification from Facebook pulls the member to Facebook, not to their plan. And the answers accumulate somewhere Boldin does not own, so a subscriber can keep the most valuable part of the relationship after cancelling. Meta owns the retention surface of a company whose entire differentiation is alignment with the end user.

The classes are the opposite case, and they are the most underrated asset here. Boldin runs live Zoom sessions most weekdays, an 8-week intro planning course, 24+ recorded courses, and weekly Q&A.6 That is already a ritual with a calendar, a cohort, and a human on the other end. If I could run one query before writing any roadmap, it would be renewal rate by number of live sessions attended in the first 90 days, because a recurring scheduled event that a member shows up to is the closest thing this product has to a habit, and if the correlation is strong the entire engagement roadmap has an obvious center.

5. Linked accounts versus manual entry is probably the cleanest retention split in the data, and a lot of users decline the link

PlannerPlus includes account linking through the major aggregators, and connected accounts refresh daily.19 But reviewers consistently note that linking is optional and that many users prefer manual updates, citing privacy: they do not want to hand brokerage credentials to a third party.911 That preference is rational, especially for an audience whose entire net worth is in the accounts in question, and it is not going to be argued away.

It also creates two structurally different products under one subscription. A linked plan is alive: balances move, the projection changes, and the product has something new to say without the member doing anything. A manual plan is a snapshot that starts decaying the moment it is saved, and the burden of keeping it true falls on the subscriber, who by definition has stopped opening the app. Every engagement mechanic in Finding 1 depends on having something new to say. Without linking, there mostly is not.

So my expectation, stated as a falsifiable prediction I would check in the first week: linked subscribers renew materially better than manual ones, and the gap is larger than the gap between any two feature-usage segments. If that is true, the retention roadmap has an unusually concrete top item, and the interesting question becomes when and how the link is asked for rather than whether to push harder. Right now the ask appears to sit in setup, which is the moment of lowest trust and highest cognitive load, before the product has demonstrated anything.

6. The free tier and Boldin AI have both moved the paywall line, and nobody appears to have redrawn it

The Basic tier is generous by design: 100+ inputs, two interactive explorers, 10+ charts, limited AI Q&A, up to five personalized alerts, the weekly newsletter, and a free discovery session with a CFP.1 That is a deliberate strategy and consistent with the mission, and it is also the thing every PlannerPlus subscription has to beat, on the same pricing page, every year.

Boldin AI complicates it further. The assistant launched in February 2026 with beta numbers that are genuinely striking for this category: 15,000+ users, 190,000+ questions, an average of 8 interactions per user per day, and 90%+ positive ratings.5 Eight interactions a day from an audience that used to visit a desktop planner occasionally is a different product. It is the most plausible engine of the planning habit, and it is simultaneously the surface most capable of answering the one question that would otherwise have justified a renewal.

The current line, as far as I can see from outside, is drawn on quantity: limited AI Q&A and five alerts on Basic, unlimited on PlannerPlus. Quantity limits are the weakest kind of paywall because they punish the user without teaching them what they are missing, and they invite the user to ration rather than upgrade. The stronger line for this product is drawn on time and consequence: one-off answers are free, but the maintained plan, the saved scenarios and side-by-side comparison, the tax projections, the ongoing monitoring, and the decision calendar are what a subscription buys. That is a line the member can explain to themselves, which is the actual test of a good paywall.

What's already working

It would be a bad read of this business to arrive with six problems and no acknowledgment of what is genuinely strong. Four things stand out, and three of them are retention assets that most subscription products would have to build from scratch.

• The alignment story is real and it is the moat. No product sales, no AUM fee, no commissions. In a category where the default alternative has a conflict baked in, "we only make money if you keep finding this useful" is both an honest positioning and the reason a subscriber will link their entire net worth. It is also a constraint that rules out the manipulative half of the retention playbook, which I would treat as a feature.
• The classroom is a ritual engine that already exists. Live sessions most weekdays, an 8-week intro course, weekly Q&A, 24+ recorded courses.6 Recurring scheduled events with a human on the other end are exactly what a "planning habit" is made of, and this is running today. Most companies asking a PM to build engagement rituals do not hand them one already staffed.
• Boldin AI's beta numbers describe a frequency change, not a feature. 8 interactions per user per day, 190,000+ questions, 90%+ positive.5 Whatever else is true, the product now has a surface people touch daily. That is the raw material of a habit, and it did not exist eighteen months ago.
• The price lock is an accumulating switching cost, granted honestly. Every year a member stays, their locked price gets better relative to the market rate.2 Very few consumer subscriptions have a retention mechanic this clean that also feels like a gift rather than a trap.

How I'd measure it

The natural 90-day deliverable is a retention metrics framework the company can align around, so this section is the closest thing to an answer to the actual first assignment. The framework I would propose has four layers, and the point of the structure is that each layer is a leading indicator of the one above it, so a team can act on Tuesday instead of waiting for a renewal event a year away.

The segments I would insist the framework carries, because a blended number for this audience is close to meaningless: distance from the retirement date (accumulating, transitioning, decumulating), linked versus manual, single versus household plan, self-serve versus enterprise-channel origin, and age relative to the decision calendar (59½, 62, 65, 67, 73). Retention for a 45-year-old modeling a house purchase and a 68-year-old sequencing withdrawals are different products with a shared codebase.

Where I'd start

Mapped to a realistic first-year set of milestones, and deliberately weighted toward queries and conversations before roadmap.

1. 1Weeks 1 to 3: the survival curve and three splits. Renewal by cohort; voluntary versus involuntary; linked versus manual; reactivation rate of lapsed subscribers by months since cancel. Those four queries decide which half of this document is worth pursuing. If involuntary churn or reactivation is large, the priorities reorder immediately.
2. 2Weeks 2 to 6: define activation empirically, and talk to the people who left. Regress candidate events against renewal, and interview 15 to 20 subscribers across the segments: renewed, lapsed, paused with a stated return date. These people tend to answer when you call, so use that. The question is not "what do you want," it is "walk me through the last time you opened Boldin and what you did next."
3. 3Weeks 4 to 8: ship the first experiment. The price-lock disclosure at renewal and cancellation. It is cheap, it is honest, it needs almost no engineering, and it produces a clean read on whether the renewal moment is a value judgment or an inattention problem. Shipping something small and measurable early also buys credibility for the larger bets.
4. 4By day 90: publish the framework. The four-layer structure above, with the segments, the validated activation definition, and the guardrails, socialized with Data, Marketing, CRM and the classroom team so retention is a shared scoreboard rather than a product-team metric.
5. 5Months 3 to 6: the decision calendar. The one structural bet, sequenced after the data says which return moments matter. Start with a single, high-confidence decision (the Roth conversion window is the obvious candidate given the audience and the product's existing strength) rather than building a general engine.

Ideas to test

Ordered by cost to run, not by expected size. Each names the finding it comes from and, where there is one, the version of it I have already run somewhere else. Items 9 through 13 come directly from walking the live signup flow on 2026-08-27, not from public pages alone.

1. 1Price-lock disclosure. Show members their locked rate and what returning would cost, on the account page and in the cancellation flow. Hypothesis: save rate rises on a true, material fact with no discounting (Finding 2).
2. 2Dunning and card-updater for annual renewals. Pre-renewal payment-method check plus a real retry sequence. Hypothesis: a measurable slice of "churn" is a dead card (Finding 2; the DIRECTV prepaid-card work is the same insight from the other direction).
3. 3Your year in review, pre-renewal. Scenarios run, assumptions changed, projected net worth delta, classes attended. Hypothesis: making a year of invisible value visible lifts renewal, with the biggest effect on medium-engagement members (Findings 1 and 2; same mechanic as the Allianz "$100,000 medical bill" concreteness pass, +3%).
4. 4Randomized intro-course invitation. Invite half of new subscribers to the 8-week class, measure M12 renewal. Hypothesis: a scheduled recurring event is the strongest existing retention lever and this separates causation from self-selection (Finding 4).
5. 5Retimed linking ask. Move the account-link prompt from setup to immediately after the first insight, reframed around keeping the projection true. Hypothesis: link rate rises without pushing on privacy, and plan freshness follows (Finding 5).
6. 6Proactive AI, one decision only. For subscribers in a Roth conversion window, have the assistant open the conversation with their own numbers instead of waiting to be asked. Hypothesis: proactive AI sessions predict renewal far better than reactive ones (Findings 1 and 6).
7. 7Painted door on the household plan. An "add your spouse or partner" option at test points in onboarding, measuring demand before any shared-access engineering. Hypothesis: household plans activate and renew better, because the decision is already being made by two people (Finding 3; the Allianz fake-door discipline that killed wedding insurance in two days for near-zero cost).
8. 8Scheduled pause instead of cancel. For members who say the plan is done, offer an explicit "pause until [Medicare eligibility / retirement date / next tax year]" with the price lock preserved. Hypothesis: a named return date converts a permanent loss into a deferred renewal, and it is a more honest offer than a discount (Finding 1).
9. 9Delay the email and account creation until after a first number. Let a visitor run a lightweight version of the projection before asking for anything, and only ask for an email to save it. Hypothesis: signup converts better when it is the price of keeping something you already have, rather than the price of finding out (Finding 3; live signup walkthrough, an email was required before any output).
10. 10User-test the onboarding's craft, not just its structure. The flow is well-sequenced, but the copy and visuals read as templated or AI-generated rather than authoritative, for an audience being asked to hand over its full financial picture. Hypothesis: a more distinctly human, higher-craft onboarding raises completion and trust scores; worth a moderated usability test if one has not already run (Finding 3; live signup walkthrough).
11. 11Replace the quiz feel with visible progress, and add "finish later." A step counter or completion bar instead of a question-by-question quiz format, plus an explicit save-and-exit path. Hypothesis: onboarding currently asks for a lot of disclosure before the subscriber has had a reason to trust the product, and making progress visible and exit safe reduces abandonment without reducing what is eventually collected (Finding 3; live signup walkthrough).
12. 12Cut or make the signup video skippable. Hypothesis: a mandatory video ahead of the actual product adds friction with no offsetting trust benefit for a self-directed audience that came in wanting the tool, not a pitch; a skip option should not hurt completion, and is worth testing against the current mandatory version (Finding 3; live signup walkthrough).
13. 13Move the credit-card ask to the end of the trial, not the start. Even on a free trial, asking for a card before anyone has seen a projection is a trust ask made before the product has earned it. Hypothesis: a truly card-free trial start raises top-of-funnel signups, and the real test is whether the resulting cohort still converts at the card-collection step it now has to pass instead (Finding 3; live signup walkthrough).
14. 14Push identity or document verification as late as possible. Any photo or document upload used for verification should be asked for only once the subscriber has a reason to keep the account, not during first setup. Hypothesis: the same request lands very differently before value is shown versus after; making it a completable-later step rather than a gate should raise completion without weakening verification (Finding 3; live signup walkthrough).

A note on scope

This was built mostly from public sources: boldin.com's pricing, classroom, coaching, community and comparison pages, the Boldin help center, the company's press releases (Series A, rebrand, AI Planner Assistant launch, the July 2025 price change), third-party software reviews, and public forum threads where subscribers discuss their own renewal behavior, plus a partial walkthrough of the live signup flow on 2026-08-27 (intro video, account creation, the onboarding sequence, and the card-on-file step, stopped short of completing setup). I did not walk the trial itself, the paywall or the cancellation flow, which is the natural next step in the same format as my JamLoop and Net32 teardowns, and Findings 2, 3 and 5 would all sharpen or change with those captures. No internal churn, cohort, billing, entitlement or experimentation data was available at any point. Every finding, priority and metric above is a hypothesis built from outside evidence, not a confirmed diagnosis, and each one's first move in the job is a question or a query, not a redesign.

Notes