CRM marketing 2026: lead scoring, funnel and return of lost customers

Once a month one of the founders writes to me: “We implemented CRM, we are entering data, but sales have not increased.” I know what they did was wrong - and it wasn't the system's choice.

I have been working with CRM marketing since 2019. During this time, I went through amoCRM, Bitrix24, Mindbox, RetailCRM on twelve projects - from e-com with a turnover of 40M rubles per month to B2B SaaS with a transaction cycle of 90 days. The numbers in the article are from practice and correspondence with colleagues, not from industry reports.

CRM without processes is Excel with a beautiful interface. Scoring, cadence, win-back are not “tricks”, but a minimum set, without which CRM works like a phone book.

1. CRM as a system vs CRM as a tool: what is the difference

Most companies use CRM as a tool to enter contacts, write comments, set tasks. This is not CRM marketing. This is a digitized notebook.

CRM as a system is when every lead has a point, every manager has a touch schedule, every segment of “departed customers” has an automatic win-back chain. When data from CRM goes to the advertising account and reduces CPL. When you know at what stage of the funnel the most money is lost - and you can fix it in 2 weeks, not in a quarter.

The difference in results: on a project in medtech, where I joined at the beginning of 2025, CRM was “implemented” for a year. Contact rate - 17%. After setting up scoring and sales cadence for 6 weeks - 38%. Not a single new channel was connected, the budget was not touched. They just changed the process.

2. Lead scoring: weight model from scratch

Lead scoring — a weighting system for assessing lead quality. The bottom line: Not every incoming lead is worth the same amount of managerial effort. Scoring allows you to prioritize automatically, without meetings.

Basic model - five parameters with weights:

Total: 0-100 points. Priority A - from 60, B - 40-59, C - below 40. Leads of category C automatically go into the nurturing chain without a manager. This does not mean that they are abandoned - it means that they are warmed up through email and retargeting to state B.

On one of the e-com projects, scoring removed 40% of leads from the managers' queue - those who ended up in category C. The conversion of the remaining A and B increased from 4.2% to 7.8% per quarter, because managers stopped wasting time on non-target negotiations.

3. BANT and its restrictions in 2026

BANT (Budget, Authority, Need, Timeline) is a classic B2B qualification that is already 60 years old. She's working. But in 2026 it is not enough.

Three real limitations I face:

Authority has become plural. Previously there was only one decision maker. Now in B2B SaaS, 3-5 people influence the transaction: CEO, CFO, CTO, champion user, buyer. Reaching out to just one is not enough; the deal will get stuck at the “approval” stage.

Budget is often "unspecified" on purpose. The lead doesn’t know the budget not because there isn’t one, but because he hasn’t received approval yet. They have a task, no numbers. To qualify this person as a B is not an A or a C. There is a real need, the money will appear in 30-60 days.

Timeline is lying. “We need it next quarter” means anything in practice. The question is not “when will you buy”, but “when is your next budget approval point.” That's another question.

How I adapt BANT in practice: I add two clarifying questions - “who else is involved in the decision?” (Authority+) and “do you already have a similar budget in another direction?” (Budget proxy). They more accurately show the real quality of the lead.

4. Sales cadence: a touch schedule that works

Sales cadence - this is not “call several times.” This is a structured schedule of touches across multiple channels at specific intervals. Without it, managers either don’t get through (one call is a waste), or they call too often and become annoying.

My workflow for B2B SaaS with Priority A:

Contact rate for such a scheme on a medical technology project is: 38%. Before implementing cadence (one call + one letter) - 17%. The difference is 2.2 times - no additional leads, no new budget.

A breakup letter on day 14 is not a failure, but a tool. Some leads respond precisely to this: “No, wait, we still need this.” This is the conversion of “silent” leads that would go nowhere without cadence.

5. Win-back: the cheapest source of revenue

Win-back — reactivation of customers who have not purchased for a long time. It is the most overlooked source of revenue in marketing. Almost all the companies I worked with did not do win-back at all - although the database included thousands of “left” ones with a purchase history.

Why win-back is more profitable than attraction:

CAC win-back on one of the e-com projects - 1 500 ₽. CAC of a new client via Yandex Direct - 15 000 ₽. The difference is 10 times. The departed client already knows the brand, already trusts it, and has already gone through the purchase process. He doesn't need to explain why he can trust you.

The win-back sequence structure I use is:

Letter 1 (day 1). Not about a discount. About the benefits - new materials, case studies, useful content. Goal: to remind about yourself without sales pressure.

Letter 2 (day 7). Personal offer. A discount or bonus tied to a previous purchase. Not “10% discount on everything,” but “20% on the category you took last time.”

Letter 3 (day 21). Last chance. Short, direct: “Your bonus expires in 3 days.” No lyrics.

Conversion to reactivation on e-com is 8-12% of the “dormant” segment. For B2B - 5-8%. These are hundreds of closed transactions from the database that was already paid for during the initial attraction.

6. Integration with advertising: how CRM data reduces CPL

This is something that most marketers don’t do, although there are tools in every advertising account.

Three working mechanics:

Look-alike from the best clients. Unload a segment of buyers with above-average LTV from CRM → upload to Y.Audience or VK Ads → build a look-alike. CPL for such an audience is 25-35% lower than for targeting standard interests. I tested it on three projects - it works stably.

Exclusion of those who purchased. Basic hygiene that half of companies don’t do: unload your customer base as an “excluded audience” in acquisition campaigns. Budget savings - 10-20% for non-targeted impressions.

Retargeting for “hot” leads. The segment “qualified, did not reach the deal in 30 days” is a hot audience for retargeting. Conversion is 3-4 times higher than cold traffic. Upload them with a separate campaign with a different creative - not “learn more”, but “compare us with competitors” or “look at the case”.

Technical implementation: most CRMs (amoCRM, Bitrix24) support exporting segments by email/phone. Ya.Audience and VK Ads accept hash loading. You can automate the unloading via n8n or Make — schedule once a week, without manual work.

7. CRM marketing metrics: what to count

The main mistake is measuring the number of leads instead of the quality of movement through the funnel. Here is a funnel with benchmarks that I use to diagnose projects:

If your contact rate is below 25%, there is most likely a problem with the lead processing speed. Leads who are called within the first 5 minutes of an application convert 9 times better than those who are called an hour later. This is not a pretty number - this is what I see in the data every time I do an audit.

Additionally, I consider: the average transaction cycle in days (the goal is to reduce by 15-20% through cadence), the cost of CAC win-back vs CAC of a new client, LTV by scoring segments (A/B/C - the difference is usually 2-3 times).

Once a quarter, be sure to recalculate the weights of the scoring model. If leads with a high score are not converting, then the weights are incorrect. This is a live model, set it and forget it.

8. What to do right now

CRM marketing is not automation for the sake of automation. These are three specific mechanics, each of which gives a measurable result: scoring removes unnecessary workload from managers, cadence doubles the contact rate, win-back returns revenue at a price 5-10 times cheaper than a new attraction.

Start with one thing: unload leads for the last 90 days, give them scores according to the model from section 2, and see how many A-leads are still without a second touch. Usually 20-30% of lost revenue is hidden there.

Related materials: LTV/CAC calculator, cold email chain templates, what is lead scoring, how to build a sales funnel.

If you want to analyze your specific situation, write to Telegram or through form. Starting consultation - 0 ₽.