Marketing Strategy / Field guide
How to Build a Marketing Plan for a New Business (Without the Fluff)
AlexPublished · Updated
A one-page marketing plan for new businesses, with worked examples: pick your first customer, write your offer, find your first two channels, calculate a test budget and track five numbers.
What a marketing plan is really for
Most new businesses don't fail at marketing because they picked the wrong channel. They fail because they tried five channels at once on a small budget, with no clear message and a website that couldn't turn visitors into enquiries.
A marketing plan is a short list of decisions: who you are going after first, what you will say, where you will say it, how much you will spend testing it, and when you will decide whether it worked. Its main job is to stop a small budget being spread so thin that nothing produces a clear answer.
Here is what that looks like with the same monthly budget:
After 3 months: four channels with a handful of enquiries each. None has enough data to keep or stop.
After 3 months: two channels with enough enquiries to judge. You keep the winner and reinvest.
enough to learn from
Illustrative example. The dashed line is the rough monthly spend one channel needs before its results mean anything; it differs by market and channel.
Step 1: Pick one clear customer
Start with the customer most likely to buy soon, not everyone who might ever buy. "Small businesses" is not a customer. "Dental clinics in Kyiv opening a second location" is: you can find them, speak to them and reach them.
Write four lines about them:
- Who: type of person or business, location, size.
- Problem: what they need solved right now that you solve.
- Trigger: what makes them start looking, such as a launch, a slow month, a new competitor, a move or a deadline.
- Where they look: Google, Instagram, a friend's recommendation, an industry group, a marketplace.
The trigger matters most. It tells you when to show up, and often where: someone who has just signed a lease searches very differently from someone who is idly browsing.
Step 2: Write an offer that makes you the obvious choice
Your message has to say who it is for, what result they get, what pain they avoid and why they should believe you. If a stranger cannot repeat what you do after reading your homepage headline, the message is not ready.
Use one sentence: "We help [customer] get [result] without [pain], because [proof]." Proof beats adjectives every time. Real prices, timelines, numbers of clients, named people and examples are believable. "High quality" and "innovative" are not.
Examples are illustrative. Notice what the strong versions add: a specific customer, a concrete detail and something checkable.
Once the sentence works, use it everywhere: homepage headline, ad text, Instagram bio, the first line of your sales emails. Consistency is what makes people remember you.
Step 3: Choose one or two channels to start
Choose the channel that matches how your customer already looks for a solution, not the one that is fashionable. Use the matrix below to shortlist, then apply these rules of thumb:
- Existing demand (people already search for it): Google Ads now, SEO for later. Ads catch the search today; SEO catches it for free in six months.
- Local and visual: Facebook and Instagram ads plus a complete Google Business Profile, to reach nearby people who aren't searching yet.
- Trust-based, slow decisions: referrals, partnerships and LinkedIn. Borrowed credibility beats reach.
- New category that needs explaining: content plus email, so people understand the problem before you pitch the solution.
Don't add a third channel until the first two have either clearly worked or clearly failed after a fair test.
- Best when
- People already search for your service
- Cost model
- Pay per click
- Need before starting
- Landing page per service, call/form tracking
- Main risk
- Broad keywords burn budget fast
- Best when
- Visual product, local or impulse buy
- Cost model
- Pay per reach
- Need before starting
- Photos/video, a clear offer, Pixel installed
- Main risk
- Creative wears out in weeks
- Best when
- Customers come to you or you serve an area
- Cost model
- Free, time only
- Need before starting
- Verified address, photos, a review routine
- Main risk
- Few reviews = invisible next to competitors
- Best when
- High trust, high price, slow decision
- Cost model
- Commission or time
- Need before starting
- A few happy clients and a simple ask
- Main risk
- Unpredictable volume
- Best when
- You sell to businesses by role
- Cost model
- Time (organic) or high CPC (ads)
- Need before starting
- Founder profile, regular posts, a clear niche
- Main risk
- Slow to build an audience
- Best when
- Steady search demand, long-term plan
- Cost model
- Time or specialist fees
- Need before starting
- A page per service, technical basics
- Main risk
- Months before first results
- Best when
- Customers need educating before buying
- Cost model
- Time
- Need before starting
- Something worth subscribing for
- Main risk
- Easy to stop before it compounds
"First results" means first enquiries you can learn from, not profit. Typical, not guaranteed.
Step 4: Set a test budget (and work it out backwards)
There is no universal "right" marketing budget, and percent-of-revenue rules don't help when you have little revenue yet. Work backwards from what a customer is worth instead:
- Customer value: what a customer pays you, over the first sale or over the time they typically stay.
- Profit per customer: value × margin. This is the most you can pay to win one and break even.
- Cost per customer: cost per enquiry ÷ the share of enquiries that become customers.
- Test budget: customers you want × enquiries per customer × cost per enquiry.
- Most you can pay per customer (its profit)200
- Actual cost per customer75
- Enquiries needed / month50
Monthly test budget750 — profitable at these numbers.
Replace the example numbers with yours. If you don’t know cost per enquiry yet, use a cautious guess and correct it after month one.
If the calculator says each customer loses money, don't just spend more. Raise the price, improve how many enquiries you close, sell something else after the first sale, or choose a cheaper channel.
Commit to the budget for two to three months. Ad platforms and SEO both need time, and judging after two weeks mostly measures noise. Here is what each month of the test is for:
- 1Month 1 · Launch & learn
DoLaunch, check tracking works, fix the landing page, read search terms/comments weekly
Don’tJudge results or change everything daily
- 2Month 2 · Improve
DoCut what clearly wastes money, test 2–3 new ads, speed up replies
Don’tAdd a new channel
- 3Month 3 · Decide
DoCompare cost per customer with your maximum: keep, fix or stop
Don’tKeep a channel “just in case”
Step 5: Make your website and follow-up ready
Marketing only creates interest. Your website and follow-up are what turn that interest into customers, and this is where most new businesses quietly lose money. Before spending anything, walk the whole path a customer takes and look for leaks.
- 1Ad or postPromises something the page doesn’t mentionRepeat the ad’s offer in the page headline
- 2Landing pageSlow on mobile; three competing buttonsOne offer, one main button, loads in under 3 s
- 3Form or callTen fields, or nobody answers the phoneName, contact, one question; call back missed calls same day
- 4NotificationLands in an inbox nobody checksTelegram or CRM, with one named person responsible
- 5First replyReply tomorrow; they already booked elsewhereReply within the hour in working time, using a template
- 6Quote → customerNo follow-up after sending a priceFollow up on day 2 and day 7; log the outcome
Test it yourself once: click your own ad, fill in your form and time the reply.
The minimum before you spend on ads:
- A landing page that repeats the exact offer from the ad, with one main button.
- A short form: name, contact and one qualifying question.
- Enquiries that arrive somewhere you actually look, such as Telegram, WhatsApp or a CRM, with a named person responsible.
- A reply time target, ideally within the hour during working hours, and a template so replying is fast.
- Two follow-ups for anyone who got a quote and went quiet.
Step 6: Measure a handful of numbers
Track five numbers per channel every month, not thirty:
- Enquiries: how many people got in touch.
- Cost per enquiry: spend ÷ enquiries.
- Close rate: customers ÷ enquiries.
- Cost per customer: spend ÷ customers. Compare it with the maximum from Step 4.
- Time to first reply: the fastest lever you control.
Then make one of three decisions per channel: keep it (cost per customer is below your maximum), fix it (close to your maximum, so improve the ad, page or follow-up), or stop it (well above after a fair test). Under about ten enquiries, you don't have enough data yet.
Here is what that looks like in practice after three months:
| Channel | Spend | Enquiries | Per enquiry | Close rate | Per customer | Decision |
|---|---|---|---|---|---|---|
| Google Ads | 750 | 38 | 20 | 18% | 107 | keep |
| 750 | 21 | 36 | 19% | 188 | fix | |
| TikTok | 300 | 6 | 50 | 0% | — | too little data |
- Google Ads: Below the 150 maximum. Keep, and move budget here.
- Instagram: Slightly above. Enquiries are fine, close rate is low: fix the follow-up first.
- TikTok: Under 10 enquiries: not a fair test. Fund properly or drop.
Fictional numbers for a business whose maximum cost per customer is 150. Build the same five columns in a spreadsheet.
Your one-page marketing plan
Bring it together in seven lines. Below is a filled-in example for a fictional clinic, so you can see the level of detail that makes a plan usable. Copy the headings into a document, fill in your own and put the decision date in your calendar. Review the numbers monthly and the whole plan every quarter, or sooner if a channel clearly isn't working.
- 1First customer
- Office workers 25–45 within 3 km of our Lviv clinic with back or neck pain; trigger: pain that hasn’t gone away in a week
- 2Offer & message
- We help office workers get rid of back pain without months of painkillers, because every plan starts with a 60-minute assessment by a certified physio
- 3Channels (max 2)
- 1. Google Ads on “physiotherapy + Lviv” searches 2. Google Business Profile + reviews
- 4Test budget & period
- 800 / month for 3 months (calculator: 40 enquiries → 10 patients)
- 5Website & follow-up
- Back-pain landing page · form to Telegram · reply within 1 hour, 9:00–20:00
- 6Numbers to track
- Enquiries, cost per enquiry, close rate, cost per patient (max 150), reply time
- 7Decision date
- 1 March: keep, fix or stop each channel
An empty line in your own plan = your next task.
A fictional physiotherapy clinic. Copy the seven headings into a doc and fill in your own.
If you can't fill in a line, that's the next thing to work on. An empty "first customer" line means more customer conversations. An empty "follow-up" line means fixing your enquiry handling before any ad spend.
Key takeaways
- A marketing plan is about focus: one customer, one message, one or two channels.
- A small budget split across many channels produces no clear answers. Fund fewer channels properly.
- Work your budget out backwards from customer value, margin and close rate, and commit for two to three months.
- Fix your landing page and lead follow-up before spending on ads. That's where most money leaks.
- Track five numbers monthly and decide keep, fix or stop for each channel on a set date.