Startup marketing strategy: channels by stage and a 90-day plan
Short answer Pick channels by how many paying customers you have, not by what's popular. Before your first ten, sell directly and claim the free listings. Then test one channel where people already look for you, like search. Add a second only after the first pays back. Track every customer's source from day one, or 90 days teach you nothing.
Want help applying this to your business? Talk it through on a free 30-minute call.
Book a callWhich marketing channels should a startup use first?
The ones that fit your stage. Count your stage in paying customers, not funding rounds: you may never raise money, and a plumber and a software founder both start at zero.
| Stage | You're here when | Spend your time on | Leave for later |
|---|---|---|---|
| 1. First customers | Fewer than about 10 paying customers | Direct outreach to people you know and their referrals, free listings, a simple site, asking every customer for a review | Brand campaigns, agencies, many social accounts |
| 2. Find a repeatable channel | Customers come, but you can't say from where | One channel where buyers already search: a small search ad test, or pages that answer buyer questions. Start an email list | A second paid channel |
| 3. Grow what pays | One channel brings customers for less than they're worth | More budget or pages in that channel, then test a second | Anything you can't track to a paying customer |
A "channel" is a route customers take to find you: Google search, Maps, referrals, email, Instagram, ChatGPT answers. The common mistake: spreading a small budget over five routes, so none gets enough to show whether it works.
How do I choose a marketing channel for my business?
Start where your buyers already go when they need you. If people search for what you sell ("emergency plumber", "bookkeeping software"), search comes first. If they don't know the product exists yet, you have to go to them: direct outreach, partners, or social.
| Channel | How fast | Cash cost | Fits when |
|---|---|---|---|
| Your network and referrals | Days | None | Always, at stage 1 |
| Google Business Profile | Days to set up | Free (Google) | You serve customers in person or in a local area (setup) |
| Search ads | Days; most ads are reviewed within one business day (Google Ads Help) | You set the budget; no minimum | People search for your service and one customer is worth more than a few clicks |
| SEO pages (showing up in free Google results) | Weeks to months | Mostly your time | Buyers research before they buy (how long it takes) |
| Email list | Weeks to build | Low | Customers buy more than once, or take time to decide (build one) |
| One social platform | Weeks | Time; ads optional | Your customers spend time there and your work photographs well |
| AI answers (ChatGPT, Gemini) | Months | Mostly your time | People ask assistants for recommendations in your field (plan) |
On timing, Google's own guide to SEO is plain:
"Some changes might take effect in a few hours, others could take several months." (Google Search Central: SEO Starter Guide)
That's why stage 2 often runs a fast channel (ads, outreach) beside a slow one (pages that answer questions): the fast one brings customers this month, the slow one keeps working after.
What should a startup's marketing budget be?
No official percentage exists for a new business. The common rules of thumb are for companies that already have revenue. Canada's Business Development Bank (BDC) gives this:
"A common rule of thumb is that B2B companies should spend between 2 and 5% of their revenue on marketing." (BDC)
For businesses selling to consumers, BDC says the share "is often higher—between 5 and 10%", and its 2019 survey of more than 1,400 Canadian businesses found average small business marketing costs of "just over $30,000 a year" (Canadian dollars).
Before revenue, a percentage of nothing is nothing. Set a test budget you can afford to lose, sized to buy enough clicks or hours to learn something. Google Ads has no floor: "There's no minimum spending commitment, and you set and control your own budget." (Google Ads Help: Google Ads: Definition). How to size an ad test: Google Ads budget for a small business.
Whatever you spend, the US Small Business Administration's test is the one that matters:
"Plan to compare your marketing and sales costs to the revenue it generates." (SBA: Marketing and sales)
Should a startup start with paid ads?
Only if you can tell which ad clicks became paying customers. Ads give fast data, but a click, a call or a booked meeting isn't a sale.
One example from our own work: across Google, Meta and ChatGPT ads, an accounting firm we run marketing for spent about $6,100 from October 1 to November 4, 2026. Those ads brought about 13 booked calls and $0 in new-client revenue. Every client who paid in that period came from ChatGPT answers or a Reddit comment, at no ad cost. One firm's result proves nothing about yours. It shows why the scoreboard is paying customers, not leads.
The 90-day startup marketing plan
Four blocks. Each ends with something you can check.
| Days | Do | Done when |
|---|---|---|
| 1–14: set up | Write down who buys, what problem sends them looking, and the words they use. The SBA's advice for this part of a plan: "Describe your audience in detail." Set up Google Analytics with your calls, forms and bookings as key events (how). Add Search Console, free per Google. Claim your Google Business Profile if you serve a local area | A test form fill shows up in Analytics, and you can describe your buyer in two sentences |
| 15–30: first customers | Contact people you know and ask for introductions. Ask every new customer "How did you hear about us?" and write it down. Ask every happy customer for a review. Add an email sign-up with consent | Every customer has a recorded source |
| 31–60: one demand channel | Pick one: a small search ad test on a few keywords that show intent to buy (first campaign), or 3–6 pages answering the questions buyers ask before they buy (plan). Tag every link you post or send with UTM tags | The channel has run long enough to bring at least a few leads, or clearly hasn't |
| 61–90: decide | For each channel, divide what you spent (money and hours) by the paying customers it brought. Compare that with what a customer is worth to you (how). Keep, fix or drop each one. Write the next 90 days | You know your one best channel, with numbers |
UTM tags are short labels added to a link so Google Analytics knows where a visit came from. Google's rule:
"When you add parameters to a URL, you should always use utm_source, utm_medium, and utm_campaign." (Google Analytics Help: Collect campaign data with custom URLs)
What rules apply to a startup's reviews and emails?
Two sets catch new businesses most often.
Reviews. Ask, but don't pay or filter. Google's policy for Maps reviews bans businesses that "Offer incentives – such as payment, discounts, free goods and/or services - in exchange for posting any review" and those that "selectively solicit positive reviews from customers" (Google Maps user-generated content policy: Fake engagement). Ask every customer, not only the happy ones.
Email. The rules differ by country.
- US: The FTC says CAN-SPAM "doesn't apply just to bulk email", and "You must honor a recipient's opt-out request within 10 business days." Each violating email can cost up to $53,088 (FTC: CAN-SPAM compliance guide).
- Canada: Canada's Anti-Spam Legislation (CASL) requires consent before you send. Section 6 prohibits a commercial message unless "the person to whom the message is sent has consented to receiving it, whether the consent is express or implied".
If you'll ever email Canadians, collect consent from everyone. One rule for the whole list is simpler.
Example
A new house cleaning business, stage 1
Zero customers and $300 a month for marketing. Days 1–30: claim the Google Business Profile, text 40 people they know, offer to clean for friends of friends, and ask each customer for a review. Days 31–60: $300 on search ads for "house cleaning" in two towns. By day 90 they have 8 customers: 5 from referrals, 2 from ads, 1 from Maps. The ads cost $150 per customer; a weekly client is worth far more, so ads stay and get a bigger budget.
A two-person software startup, stage 2
20 paying customers, nobody sure where they came from. They add a "How did you hear about us?" field at sign-up and UTM tags on every link. By day 60, most new sign-ups name a comparison page they wrote or a partner's newsletter. They drop the social ads ($1,000 spent, no paying customer) and write five more pages on the questions buyers asked in sales calls.
FAQ
Which startup marketing strategy works best?
Focus. One or two channels where your buyers already look, tracked to paying customers, judged after 90 days. The best channel differs by business: a local trade often wins on Maps, a software product on search and partners.
How long does startup marketing take to work?
Outreach and ads can bring customers within weeks. Pages aimed at free Google results usually take months; Google says changes "could take several months". Plan 90 days per test, then decide.
Should a startup hire a marketing agency?
Usually not before you know which channel works, since you'd be paying someone to guess with your money. Once one channel pays back and you lack time to grow it, an agency or freelancer can make sense (how to choose).
What should a startup measure?
Cost per paying customer for each channel, and what a customer is worth over time. Followers, views and clicks only matter if they lead there (traffic sources in Analytics).
Different for you?
- You can't tell which channel brought a customer: fix tracking first (connect your marketing data).
- People find you on ChatGPT or other assistants: track AI referral traffic.
- Your buyers search on Google before they call: SEO is your slow channel; Paid Ads your fast one.
- Visitors arrive but don't call or book: the problem is the page, not the channel (Conversion, Website).
- You want content as your main channel: content marketing for startups.