StrategyJune 2, 20269 min

Subscription and MRR 2026: how to transfer customers from one-time payments to a retainer

How can a marketer or agency switch to MRR: three types of subscription, trial offer for database migration, onboarding for up to 30 days, churn and NRR metrics. The churn rate for a B2B retainer is up to 3% per month. NRR is above 100% - the base is growing without new clients.

Article cover:Subscription and MRR 2026: how to transfer customers from one-time payments to a retainer

One-time sales mean starting from scratch every month. Subscription means waking up with MRR, which is already there before the first bell rings.

I worked as CMO of EZ KATKA in 2023-2024. The network reached 19 arenas and a customer base of 120 thousand people; its turnover for 2024 amounted to 124 million rubles. This is the context of the business, not the MRR or the result of converting customers to a subscription. Below I discuss the subscription model itself and ways to calculate its economics.

Switching to a subscription is not a change in price. This is a change in business model. Price is the last thing that needs to be changed.

1. Three signs you need a subscription

The first sign is that revenue is jumping. January is good, February is bad, March is good again. The team is confused whether to plan a hire or not. The advertising budget is either cut or restored. This is not seasonality or “the market” - this is the architecture of a business without recurring income.

The second sign is that CAC is growing, but revenue is not growing proportionally. You pay again for each new client, although the old client could have simply continued paying. With one-time sales, every month is a new auction for attention. When you subscribe, the client is already yours until he leaves.

The third sign is that clients return on their own. If 30–40% of clients return after 2–6 months, then there is value. The only question is, why wait until they remember about you - you can simply fix the relationship through a subscription.

COLIZEUM for 01.01–14.12.2023 had a total coverage of VK community publications of 82.4 million with an ER of 2.6%. This data shows the extent of communication, but does not in itself describe customer retention and repeat visit costs. Subscription decisions require separate data on cohorts, margins and repeat purchases.

2. Three recurring revenue models: retainer, membership, SaaS shooting range

Not every subscription is the same. The choice of model depends on the type of business and what the client “gets” each month.

ModelWho is it suitable for?Average bill of the Russian Federation 2026Norm churn/monthMain risk
RetainerAgencies, consultants, freelancers15 000–80 000 ₽up to 3%Burnout with a large number of clients
MembershipCommunity, EdTech, clubs990–9 900 ₽up to 6%The value must be updated every month
SaaS shooting gallery (Lite)Digital Products, B2C SaaS490–3 900 ₽up to 7%High support volume with low ARPU
SaaS shooting range (Pro/Business)B2B SaaS, tools for teams5 000–40 000 ₽up to 3%Long sales cycle, high CAC
Hybrid (retainer + membership)Expert projects, courses + consulting10 000–50 000 ₽up to 4%Difficult to standardize

A retainer is the easiest model to start with. You take your current service and repackage it into a monthly package with a fixed volume and a fixed price. The client knows what he is getting. You know what you're getting.

Membership only works if the content or community is actually updated. I've seen projects where they sold a membership, recorded 10 lessons and forgot about it. After three months, churn was 25–30%. This is not a subscription - it is an installment plan for a one-time product.

SaaS shooting galleries require a product with a real differentiation of functions between tariffs. If free and paid shooting galleries differ only in the request limit, this is weak differentiation. Strong - when a paid shooting gallery provides functions without which the work slows down by half.

3. How to migrate your current database without churn

The main mistake during the transition is to announce “we now have a subscription” without preparing the client. A one-time customer views subscription as an obligation rather than a convenience.

A migration offer should remove three objections: “why should I pay every month”, “what if the price increases” and “what if I want to stop”. A simple structure that works:

  • The first month is at the old price (or with a symbolic discount of 10–15%)
  • The price is fixed for 6 months without changes
  • Exit - at any time without penalties (this removes the fear of obligation)

Migration timeline for a base of 30 clients: month 1 - mailing with an explanation of the new model and an offer for early migration; month 2 - calls or voicemails to those who did not answer; month 3 is the last chance to fix the old price. Realistic conversion: 30–40% of the base per quarter. Not 100%, and that's okay.

On one of the projects, I migrated 12 clients out of 28. It seemed not enough. But 12 × 25,000 ₽ = 300K ₽ MRR with zero CAC. Before the transition, the same income required 4–5 new projects every month. The difference is stress and predictability.

About the mechanics of customer retention when changing models - more details in the article about loyalty programs.

4. Onboarding funnel: the first 30 days determine the churn

I analyzed churn for three projects with a subscription model. There is only one pattern: 60–70% of all canceled subscriptions were from clients who did not receive a specific result in the first month. We didn’t “communicate”, we didn’t “get to know each other” – that’s exactly the result.

Onboarding structure that reduces early churn:

Day 1 - a welcome message with three specific things: what the client will receive in the first month, how to contact you, what is needed from him right now. Without “welcome,” without “welcome to our family.”

Days 3–5 – first visible result or check-in. For the retainer, this may be an interim report on the first tasks. For SaaS - an automatic letter “you have already done X actions, this is what it did.”

Day 14 - interim report with numbers. Not “we are working on your project”, but “did X, result Y, next step Z.”

Days 28–30 – first month review. Live call or voice. Ask: what did you like, what did you expect differently, what result do you want for the second month? This call increases retention in the third month by about 35–40% - at least that’s what happened on my projects.

5. MRR, NRR, Churn Rate, CAC Payback - how to calculate and benchmarks

Four metrics without which a subscription model is just “money comes in every month” and not a managed business.

MRR = sum of all active monthly subscriptions. If the client pays once a year, we divide by 12. MRR 300K ₽ with retainer 25,000 ₽ = 12 clients. Simple math, but it shows how far you are from your goal.

Churn Rate = departed clients / clients at the beginning of the month × 100%. Churn 5% per month is 46% of the base for the year. Many people don't think so because "only one or two clients left." Count in percentages, not in people.

NRR (Net Revenue Retention) = (MRR start + expansion MRR − churned MRR − contraction MRR) / MRR start × 100%. NRR is above 100% - the base grows itself. NRR is below 90% - the base is shrinking even if there are new clients. This is the most honest metric of the health of a subscription business.

CAC Payback Period = CAC / (ARPU × margin). The norm for a B2B retainer is up to 3 months. For SaaS – up to 12–18 months with NRR above 110%.

NicheNorm churn/monthNRR rateCAC PaybackRed zone churn
B2B retainer (agency)1–3%105–120%1–3 monthsabove 7%
B2B SaaS (team plans)1–3%110–130%9–18 monthsabove 5%
B2C SaaS/applications3–7%90–105%3–9 monthsabove 12%
EdTech/Courses (Membership)4–7%95–110%2–6 monthsabove 10%
Community/club3–6%95–115%1–4 monthsabove 10%

These figures are the median for the Russian Federation 2026 based on open data and correspondence with colleagues. The actual turnaround depends heavily on how well the onboarding is built and whether the client has clear value every month.

Read more about calculating CAC and LTV in the article about unit economy.

6. How to reduce churn: triggers, win-back and upgrade

Churn is reduced not by promotions or discounts. They reduce it with value at the right time.

The three peak churn times in a subscription model are the end of the first month, the three-month mark, and the annual renewal date. Each point requires a separate script.

On the 30th day, live contact is useful: a call, voice or personal message instead of a template letter from the “support team”. For a subscription model, this step should be checked by cohort: how many clients remained active after the first month and how much time the team spent on contact.

For three months - upsell or bonus. The client has survived the initial interest and entered the “middle phase” - when it still works, but is no longer as exciting. Offer an upgrade with new value: additional session, new report, expanded access. Upselling at this moment is perceived better than a month after the start.

7–14 days before the annual renewal - a letter with the results of the year and a proposal to fix the price for the next year. A client who sees concrete results within 12 months makes the decision to renew 2–3 times easier.

Win-back for those who left: a letter 30-60 days after cancellation with a simple question - “what could we have done better?” Not with an offer to return. Just a question. About 15-20% of those who leave respond, and about a third of them return within a quarter—often at a higher rate.

7. Subscription pricing: range anchoring and annual plans

One tariff is the easiest to set up and the worst to sell. Three tariffs are standard. The client chooses between your tariffs, and not between you and a competitor.

Anchoring principle: an expensive tariff makes the average one reasonable. If the average tariff costs 15,000 ₽, and the expensive one costs 35,000 ₽, the client sees “only twice as expensive for premium” - and either takes the average one as a “normal choice”, or takes the expensive one, because “the difference is not that big.” Both scenarios are beneficial to you.

A cheap anchor tariff is needed not for sales, but to reduce the barrier to entry. It should actually work, but have one significant limitation - for example, there is no strategy session or no priority response. If a cheap tariff is too good, everyone takes it, and ARPU falls.

Annual plans are the main tool for reducing churn. A client who has paid for a year, churns 4–5 times less often than monthly. The standard discount for a year is 15–20%. Math: you lose 15% of revenue, but get 4-5 times higher retention and predictable cash flow for 12 months in advance. It's a good deal.

Do not give a discount for a year of more than 25% - the client begins to doubt the value of the product (“why is it so cheap”). 15–20% is a psychologically “fair” discount for prepayment.

About pricing strategies in detail - in the article about pricing without churn.

8. Case: how I reached MRR 300K+ in six months with a base of 30 clients

At the end of 2022, I was working on one-time projects. Good money, but every month from scratch. October was a disaster - two projects were postponed. November was excellent - three closed at the same time. There is zero predictability.

In January 2023, I switched to a retainer model. The base at that time was about 30 clients with whom I worked at different times. Not all are active, but all are “warm”.

Step one. I packed three tariffs: basic retainer (15,000 ₽/month - maintaining one channel + weekly report), standard (28,000 ₽/month - comprehensive SMM + content plan + analytics), pro (55,000 ₽/month - everything from the standard + strategic session once a month + priority response).

Step two. I sent out a simple letter to 30 clients: “I’m switching to a retainer model. The first month is at the old price. The price is fixed for 6 months. Exit at any time.” No long explanations, no “unique offer”.

Results for the first three months: 11 out of 30 switched to a retainer. Distribution: 4 basic, 5 standard, 2 pro. MRR: 4 × 15,000 + 5 × 28,000 + 2 × 55,000 = 310,000 ₽.

In six months, I connected 3 more new clients through recommendations. MRR increased to RUB 390,000. CAC for new clients is almost zero (referral income). To get 390K in one-time projects, you had to close 5-7 projects per month. Now it is enough to retain 14 clients and add 1-2 new ones per quarter.

Churn for six months - 2 clients (both basic tariffs). Churn Rate 1.4% per month on average. One returned after four months to the standard rate.

What didn’t work: I tried to convert “cold” clients from a database that was three years old. Out of 8 attempts - zero. The retainer model works with those who have fresh interaction experience and an understanding of value. With a cold base, another warm-up cycle is needed before the subscription offer.

If you want to discuss the transition to a retainer model for your type of business, write to Telegram @dipustovalov or through form. I’ll look specifically at what model to choose, how to package the offer, what MRR it’s realistic to achieve in a quarter. Starting consultation - 0 ₽.

Related materials: unit economy for marketers, pricing without churn, loyalty program, CRM Marketing and Automation, Product-Led Growth.

More on the topic