The service business marketing playbook.

Why tactics without strategy fail

There is a new must-try tactic every week, and each one works for somebody. The trouble is that a tactic chosen before the strategy exists has no way to be judged. Was the campaign bad, or was it pointed at the wrong people with the wrong message? You cannot tell, so you try the next one, and the budget goes with it.

The seven common causes of failure all trace back here: no clear audience, no differentiation, tactics before strategy, inconsistent execution, nothing tracked, impatience, and no defined customer journey. Fix the first three on paper and the other four become manageable.

The three decisions

Who you are for: not everyone, the specific customers most likely to buy, pay well, stay and refer. Why you: the reason they should choose you over the alternatives, in one sentence a stranger could repeat. Where next customers come from: the two or three channels that fit those people and your economics, in order of priority.

Write each one on a page. If a decision cannot be written plainly, it has not been made yet, and the marketing built on top of it will show it.

How long, and who does it

A strategy pass is a short project: a look at where customers actually come from, what the numbers say about each channel, and where leads leak, then the three decisions and a plan you can run with the people and budget you have. Agencies sell execution. Strategy makes sure the execution is pointed at the right thing, whoever does it.

The ten-second test

Hand your website to someone who does not know your business and give them ten seconds. Then ask: what do they do, who is it for, and why would I pick them over the other options? If the answers are vague, every marketing dollar you spend is working twice as hard as it should, because the first job of each ad, page and post is explaining what you are.

Why narrow beats broad

"Retirement planning for physicians in [your region]" gets referred, searched for and remembered. "Comprehensive wealth management for individuals and families" does not, because it describes a thousand firms. Owners resist narrowing because it feels like turning work away. In practice the narrow promise brings more of the right work, and the rest still comes through the door.

Where the sentence goes

At the top of the home page, in the Google Business Profile description, in the LinkedIn headline, at the top of every proposal, and in the two-line answer when someone asks what you do. The messaging framework underneath it (the proof, the objections, the words you use for each service) keeps the website, the profile and the owner from describing three different businesses.

A website with one job

A service business website is not a brochure or a portfolio. Its job is to turn the right visitor into a conversation: a headline a stranger understands in ten seconds, one clear next step repeated where people are ready for it, proof near every ask, a contact form that takes under two minutes, and a site that is fast on a phone. Measure form fills and calls, not traffic.

The Google Business Profile

For most local service businesses the profile is seen more than the website, and it decides who gets the call from the map. Finish every field: the most specific primary category, every service, hours including holidays, a plain description, twenty real photos. Post something weekly, answer every review and question, and keep the name, address and phone identical everywhere. Fifteen minutes a week keeps it alive.

Reviews as a system

Volume and recency beat a perfect score: a business with 200 reviews at 4.7 wins over one with 14 at 5.0 almost every time. The businesses with 200 did not get lucky. They built a system: a trigger at the moment the job is done, a personal ask, a direct link, a reply to every review within a week, and a place to reuse the best ones. Automate the ask once the manual version works.

Tracking, from day one

Ask every new lead how they heard about you and record the answer. Tag every link you control with UTM parameters. Use a tracking number for the phone on each channel you pay for. Once a month, connect closed revenue back to its source. Four habits, no data team, and the difference between a budget set by feel and one set by evidence.

What still matters

Relevance, distance and prominence decide the map. Clear pages decide the rest: one page per service, one per area you serve, each saying in plain words what you do there and for whom, with proof and a way to act. Speed matters more than most owners think; Google's own research found over half of mobile visits abandoned when a page took more than three seconds. Fix the boring reasons: huge photos, too many plugins.

What stopped working

Generic blog posts nobody asked for. Keyword stuffing. Bought links. Chasing every new tool that sells an AI visibility score. Write fewer pieces that answer the actual questions your customers ask, better than what is out there, and keep them current.

Showing up in the answer

A growing share of customers ask ChatGPT, Perplexity or Google's AI Overview for a recommendation and never click anything. An AI names the clearest, most corroborated option: plain pages that state what, for whom and where; identical name, address and details across your site, profile and directories; a steady stream of real reviews and mentions elsewhere; and answers, in writing, to the questions people ask. This site's answers hub exists for exactly that reason.

A realistic timeline

Months one to two: profile complete, reviews flowing, technical fixes done; map visibility improves first. Months three to six: service and area pages published and refined; rankings for specific services in specific places start to move. Months six to twelve: reviews compound, local links accumulate, the site becomes the obvious answer for its niche. Plan for it rather than around it.

The math before the ads

A lead is worth average customer value times close rate. If that clears the cost per lead with room to spare, run ads, starting with Local Services Ads if you qualify (pay per lead, and the Google Guaranteed badge does the trust work), then one tight search campaign for your highest-value services. If it does not clear, no amount of optimizing will save the campaign; raise the close rate or the customer value first.

Five mistakes that waste most budgets

Broad match on everything. Sending clicks to the home page instead of a landing page. No negative keywords, so you pay for "free" and "jobs". Judging by clicks instead of cost per customer. And the biggest: leads that reach voicemail. A campaign is only as good as the first hour after the lead arrives.

Meta, honestly

Facebook and Instagram ads are bad at catching someone the moment they need you and good at two jobs: staying in front of people who already know you (retargeting website visitors and video viewers, always on, small daily budget) and putting one useful piece of local content in front of a tight radius at the right season. Run them from the ads manager, never the boost button, and judge them by calls and forms.

When to say no

Geofencing routine services. Any awareness campaign with no way to trace a customer back to it. Any lead marketplace that sells the same inquiry to three competitors, unless your reply is the fastest. Any channel that cannot answer, at the end of a quarter, what a customer from it cost.

Email, still

Email returns roughly $36 for every $1 spent by Litmus's estimate, because you own the list and the cost to send is close to nothing. Start with a welcome sequence and a monthly note worth opening, segment once the list grows (customers versus prospects is enough at first), and measure revenue per email rather than open rates. A list that only hears from you when you need money stops opening.

A content rhythm you can keep

Most small business content dies in March because every piece starts from zero. Start every one with a real question a customer asked, answer it properly once, and cut that answer into six outputs: a page or post, an email, a few social updates, a short video, a Google profile post and a line for sales conversations. One source a week is a month of marketing.

Thought leadership without the cringe

You do not need a personal brand. You need to be the person people trust on one subject. Pick the narrow thing you know better than anyone nearby, write short notes from real work rather than manifestos, put them where your customers already are (usually LinkedIn, email and your own site), and give it a year. Leave out the humblebrags.

The first hour

Answering a lead within the first hour, and ideally within five minutes, can double the results of every channel you already pay for, because most inquiries go to whoever responds first. The research found firms replying within an hour were about seven times as likely to qualify a lead as firms that waited an hour longer, and nearly a quarter of companies never replied at all. The system has five pieces: an instant acknowledgement, an alert to a phone, one named owner per shift, a short real reply, and a record of every lead with its source.

Capture, reply, prove, return

Capture: every call, form and message into one list with its source. Reply: the first-hour system. Prove: the review ask at the moment the job closes, with a direct link and a reply to every review. Return: a referral ask after every good outcome, and a reason for last year's customers to come back before they forget you (maintenance reminders, recall, a seasonal note). Four pieces, in that order.

Tools come last

Most businesses that buy automation platforms use a fraction of them because nobody designed the process first. Design the process on paper, run it by hand for a month, then set up the smallest amount of software that runs it reliably. Every system gets documented in plain language: the trigger, the steps, who owns it, and what to check monthly.

How much

Before setting a number, answer three questions: who exactly you are trying to reach, why they should choose you, and what the path from stranger to customer looks like. Then pick a percentage of revenue as a starting point (5% to hold steady, more to grow) and remember that an owner who bills $150 an hour and spends five hours a week on marketing is already spending about $3,000 a month before the first ad runs.

Where it goes

Foundation (about 20% until it is finished, then 8% for upkeep): the profile, reviews, tracking and website. Proven (the rest up to 70%): whatever you can already trace customers to; spend here to grow, not to experiment. Promising (20%): channels that fit but are not proven, given two quarters and a defined result. Tests (10%): one new thing at a time, and the bucket never grows. Under about $2,500 a month, collapse to foundation and proven.

The monthly review

Once a month, one table: each channel, what it cost including hours, the leads and customers it produced, and what a customer cost. Nothing moves to proven without a traced customer. Nothing stays in promising for more than two quarters. Impressions, likes and "brand awareness" do not appear on the table.

Weeks one and two: decide and finish

Write the three strategy decisions and the positioning sentence. Finish the Google Business Profile and turn on the review ask. Add "How did you hear about us?" everywhere. Build the first-hour reply system by hand: who owns the inbox, what the acknowledgement says, what the reply looks like. Score the profile and do the top three fixes.

Month one: the proven channel

Put the budget into the one channel you can already trace customers to and run it properly: if that is search, Local Services Ads plus one tight campaign with a landing page and a tracked number; if it is referrals, the ask and the thank-you as a system. Rewrite the home page for the ten-second test. Start the customer list and send the first monthly note.

Months two and three: build the slow lane

Publish one page per service and per area. Answer the ten questions customers ask most, on your own site, in writing. Set up retargeting. Pick your subject and publish the first note. Hold the first monthly review with the attribution table, and move channels between buckets based on what it says, not on how they felt.

What to expect

By the end of the quarter the follow-up system and the profile should already show in the numbers. Paid search should have a cost per customer you can stand behind or a decision to stop. The slow lane will look like it is doing nothing; that is what compounding looks like from the front. Keep the rhythm for a year and the phone starts ringing from places you do not pay for.