The first 90 days of marketing in a startup: plan before and after the first clients 2026

Most startups start marketing from the wrong place. They launch advertising before the first sales, hire an SMM specialist before they understand who the client is, and wonder why the CPL is 8,000 rubles when the conversion is zero. Below I discuss not “what is go-to-market”, but specifically: what to do in the first 30, 60 and 90 days, what not to touch at all, and where a startup loses money the fastest.

At EZ KATKA, I led the marketing of a network that reached 19 arenas and a customer base of 120 thousand; The turnover of the entire network for 2024 amounted to 124 million rubles. The summary specifically states a 15% reduction in CAC. In NEMIFIST, the Telegram channel grew to 50 thousand subscribers, and the Senler funnel collected 1,022 leads at 12.56 rubles. The scale of the business and the marketing result are shown separately here.

The first 90 days at a startup are not about marketing. This is about understanding who needs the product and why before you start spending money on it.

1. Why the first 90 days at a startup are unlike anything you've ever experienced before.

If you come from an agency or a large brand, forget about the usual processes for the first three months. In an agency, you already have a budget, a product, at least some audience and historical data on channels. A startup has none of this. There is no data for optimization, no social proof, no understanding of real objections, no proven offer.

I myself have moved from hiring to a startup environment several times - and each time the first few weeks were spent on stopping doing “right” and starting doing “fast and verifiable”. In a startup, the only truth is the market's reaction to your hypotheses. A beautiful strategy, a media plan for the year and a brand book have no meaning until you know who your client is.

Another feature: there are no resources. I've worked in situations where all the marketing was me plus half an hour a day from the founder. In this mode, “integrated approach” and “360-degree strategy” are not tools, they are excuses for those who are afraid to make one specific move and check the result. The first 90 days are a series of quick checks, not a grand strategy.

2. Phase 0-30: what to do before the first sale and what not to touch

Before the first sale there are only three things: ICP, customization, manual sales. No advertising, no SEO, no automation. Not because these are bad tools, but because without a proven ICP and offer, they work in vain.

ICP is not a “target audience.” This is a specific person in a specific company with a specific pain right now. “Small service business” is not an ICP. ICP - “the director of a 300-600 sq.m. fitness club who loses clients after the first month and does not understand why.” The more specific you are, the higher the conversion at all subsequent stages.

Castdev - 5-10 interviews with potential clients. Not a sale, but a conversation: what a pain, how they decide now, how much it costs, what they tried. The main question is: “what was the last straw before you decided to look for a solution?” The answer to it is the trigger for the offer and advertising message.

In the first 30 days of a new project, check whether there is demand, who makes the purchasing decision and how each application is recorded. SEO, newsletters and complex CRM can wait until a stable funnel appears. Don’t call outreach “zero CAC”: labor time and tools also cost money.

3. How to find your first 10 clients without an advertising budget

TG outreach is the fastest channel. The first response can be received on the day of sending. A message structure that works: context (where you know the recipient from or why specifically to him), a specific observation about his situation, value-prop in one sentence, a CTA with a minimum barrier - “I can show you what it looks like on a similar project.”

For outreach, it is useful to maintain a simple funnel: sent personal messages → responses → calls → agreements. Set your targets early and compare them to your channel, not a universal “norm.” If you attract bloggers, consider the agreements and results of publications separately from applications in the bot.

Personal recommendations work faster than outreach. Ask everyone you know in the industry: “Who in your circle needs what I do?” A referral lead closes 3-4 times faster than a cold one - it comes with trust in the sender. In this case, a landing page is not needed at all - a one-page description in Google Doc or Notion is enough.

Read more about the path from zero to the first hundred clients in the article about first 100 clients 2026.

4. Phase 30-60: after the first sales - what to measure and how not to break what works

After the first 10 deals, you have the most valuable asset a marketer can have: real customer data. Now you need to document what worked before you change anything.

Three questions that need answers by the end of the first month: where the best clients came from (not all, but with the highest LTV), what objection was encountered most often and how you closed it, what offer wording aroused the most interest. These are the three key parameters for building your first scalable channel.

What not to do on days 31-60: do not rewrite the landing page completely, do not change the offer without data, do not launch three channels at the same time. I have seen several startups that at this stage began to “optimize” what was already working - they changed the landing page because it was “ugly”, they rewrote the script because “it doesn’t sound premium.” The result was that they broke the working mechanics and lost a week or two to return to their original state.

On days 31-60, you need to set up a measurable funnel: UTM markup, Yandex Metrica with goals for key events, CRM with stages. Without this, the transition to paid traffic will be blind. How to build a funnel from scratch - details in the article about building a sales funnel 2026.

5. Mistakes that startups make on days 60-90: scaling to product-market fit

This is the most expensive mistake that I have seen five times in my practice. The team closed the first 10-15 deals, got inspired, decided “now let’s scale it up” - and launched Yandex Direct with a budget of 150-300K rubles. CPL comes out to 6,000-10,000 rubles, conversion to purchase is 0.5%, the team is disappointed in the channel. The problem is not the channel - the problem is that the PMF has not yet been achieved.

PMF—product-market fit—is not “we like the product.” Sean Ellis Metric: Ask customers, “How would you feel if this product disappeared?” PMF = 40%+ answer “I would be very upset.” If it is lower, the product does not solve the pain well enough, or the pain is not sharp enough. No amount of advertising will fix this.

The second PMF signal is a re-buy. If out of the first 10 clients at least 2-3 come back for the next month/transaction, this is a working product. If all 10 left after the first contact, look for a problem in onboarding or in matching expectations and reality. Launching a budget before this point means attracting customers who will leave in a month.

About product-led growth and how to integrate marketing into product growth - a separate topic in the article about PLG 2026.

6. Phase 60-90: when is it time to turn on paid channels and how to choose the first one

Signals for switching to paid traffic: there is PMF (40%+ “will be very upset”), there is a measurable funnel, there is at least one channel with an understandable CPL from manual sales. Without these three conditions, it’s too early.

The choice of channel depends on the type of demand. Search is useful for generated demand; for the unformed, you can test VK Ads or TG Ads. In COLIZEUM, the total coverage of VK community publications was 82.4 million with an ER of 2.6%; the available presentation does not highlight the paid share of this coverage, so it is impossible to use the number as a result of coverage advertising.

The first paid channel is tested with a minimum budget - sufficient for statistics. For Direct this is 60-100K rubles per month, less makes no sense. A test of 2 weeks for 15K rubles will not give any conclusions other than “we don’t know.” Either a full test or don’t waste your money. Read more about channel selection and GTM in the article GTM playbook 2026.

7. Metrics for the first three months: what should a marketer in a startup actually consider?

The main mistake in metrics at the start is to count reach, impressions and subscribers instead of CPL, CAC and LTV. Reach means nothing without conversion into money. I've seen startups with tens of thousands of followers on Instagram (where there isn't one) and zero sales.

Unit economics is the only metric that truly shows the health of startup marketing. About calculating LTV, CAC and unit economy for different models - read more in the article about unit economy 2026.

8. Checklist: what should be ready by the end of the 90th day

By day 30 you should have:

• ICP on one page - job title, company type, pain point, purchase trigger
• 10 paying clients from manual outreach
• Recorded source and objections for each transaction
• An offer in one sentence that has been tested on real buyers
• Minimum landing page - Tilda or even Google Doc with description

By day 60 you should have:

• First PMF measurement using the Ellis method
• Measurable funnel: UTM, Yandex Metrica with goals, CRM with stages
• Selected one scalable channel from the first 30 days of data
• First data on CPL and CR lead→deal in the selected channel
• A clear picture from the churn of the first cohort

By day 90 you should have:

• CAC on the main channel is stable, not “about”
• LTV of the first cohort - at least an estimate based on 60-day data
• Solution: increase the channel budget (LTV/CAC above 3) or pivot
• If we increase it, a media plan for the next quarter with one main channel
• If you pivot - a new ICP hypothesis and a return to manual sales

90 days is not the end, it’s the first full cycle. Either you have data for scaling or data for the next hypothesis. Both outcomes are normal. It’s crazy to spend three months on a “strategy” without a single proven sale.

If you want to review the first 90 days for a specific product, write to Telegram or through form. Starting consultation - 0 ₽.

Related materials: first 100 clients without budget, GTM playbook 2026, building a sales funnel, unit economy for a startup, product-led growth 2026.