Marketing

The Marketing Funnel and Sales Process for Service Companies

A diagram of the marketing funnel feeding the sales cycle, with the handoff between them marked as the leak

I drew this graph in 2021 for a landscaper who kept saying he had "a marketing problem." He did not. He had a nine-day gap between a form fill and a phone call. The graph is still right. What it never showed is where the money actually leaks out of it, and what the picture looks like when nobody picks up the phone at all.

So this is the 2026 version. Same spine, three additions: the handoff between the two halves, what changes when you sell a product or a digital service instead of scheduled labor, and two prompts you can paste into an AI to work through your own version out loud.

The whole picture on one line

Marketing and sales are not two departments. They are one conveyor belt with a seam in the middle, and most service companies are losing jobs at the seam rather than at either end.

Marketing: strangers become leads

AttractThey find out you exist
ConsiderThey believe you can fix it
ConvertThey raise their hand
LEADA name and a way to reach them
The handoffMost of the loss happens in the minutes right here

Sales: leads become customers

ConnectReach them, qualify them
DiscoverFind the real problem
PresentShow the fix, priced
CloseSignature or deposit
Referral loop: happy customers feed the top of the funnel. This is the only part of the picture that compounds.

Everything below is one of those boxes, what it means for a service business, and the specific way it usually breaks.

Part one: the marketing funnel

Three stages, and the only job of each is to earn the next one.

Attract: turn strangers into visitors

You are trying to be findable at the moment someone has the problem. For a local service business that moment is usually a break, not a plan. Housecall Pro surveyed 1,100 US homeowners in February 2026 and found 58 percent contact a pro because of a sudden problem and 51 percent because something is gradually getting worse, against 38 percent for a planned upgrade. (Housecall Pro, 2026 Home Services Report)

Which means attraction is mostly about being present in the places people check when something breaks. BrightLocal's 2026 survey of 1,002 US consumers puts Google first at 71 percent, Facebook second, and AI tools third at 45 percent, up from 6 percent the year before. (BrightLocal, Local Consumer Review Survey 2026)

Consider: turn visitors into believers

This is the stage everyone skips. The prospect is not comparing prices yet. They are deciding whether you are the kind of company that handles their kind of problem. Show them you have seen this exact job before.

The proof they actually look at is reviews. Ninety-seven percent of consumers read them, and they read across an average of six different review sites. Forty-nine percent say they trust online reviews as much as a personal recommendation. (BrightLocal, 2026)

Convert: turn believers into leads

Give them an obvious reason to raise their hand and an easy way to do it. Note how narrow this gate is in practice: after reading positive reviews, 54 percent of people visit the website, but only 20 percent contact the business and 20 percent book. (BrightLocal, 2026)

The quiet math of the funnel

Roughly half of the people who like what they read go look at your site. Roughly a fifth of those raise a hand. Doubling the top of your funnel doubles your ad bill. Fixing the Consider stage costs nothing and moves the same number.

The handoff, which is where the money goes

The 2021 graph has one arrow between the funnel and the sales cycle. That arrow is the most expensive inch of the whole diagram.

In 2011 Harvard Business Review published an audit of lead response times across US companies and found that firms that tried to make contact within an hour of an inquiry were around seven times more likely to have a meaningful conversation with a decision maker than those that waited even one hour longer, and more than sixty times more likely than those that waited a day or more. (Oldroyd, McElheran and Elkington, Harvard Business Review, March 2011)

That study is fifteen years old and the underlying behavior has not softened. What has changed is the competition for that hour. Sixty-eight percent of homeowners now get multiple quotes before deciding. (Housecall Pro, 2026)

Worth being precise about

You will see a widely repeated claim that "78 percent of customers hire the first business that responds." I could not trace that number to an original study, and it sits awkwardly next to a census-balanced survey showing 68 percent collect multiple quotes. The honest version is less dramatic and more useful: being first gets you into the comparison, it does not end it.

So the handoff has two jobs, not one. Answer fast enough to be in the set. Then be good enough in the next twenty minutes to win the set.

Part two: the sales cycle

Eight stages in the original graph. Here is what each one is for, and the specific way it usually fails.

1Activity

Everything you do to start conversations: the callbacks, the follow-ups, the quote reminders, the check-in on last spring's estimate.

Fails when it only happens on slow weeks, so the pipeline arrives in waves.

2Connect

Actually reach a human and find out whether this is a fit: budget, authority, need, timeline.

Fails when you treat it as a scheduling call. A connect call with no qualifying question is just a calendar invite.

3Pipeline

The queue of qualified leads. A real list with a next action and a date against each name.

Fails when it lives in a truck console and a text thread.

4Discovery

Ask enough questions that the customer describes the problem, the cost of leaving it, and what a fix is worth. You are not presenting yet.

Fails when you quote in the first five minutes. Now you are a number on a page next to two other numbers.

5Prospect

The shortlist: qualified, with a defined need you can actually solve.

Fails when everyone is a prospect. If nothing is ever disqualified, the stage means nothing.

6Present

Show how your solution fits the problem they just described, in their words, at a price.

Fails when it is the same PDF for everybody.

7Close

Make committing easy and obvious. What happens next, when, and how do they say yes right now.

Fails when the yes requires a printer.

8Referral

A system, not a hope. Ask at the moment the work is visibly finished and the customer is standing there happy.

Fails when it is never asked for, which is most of the time.

The one rule inside the cycle

Do not present solutions until the customer has told you what the problem costs them. Discovery is not small talk before the quote. Discovery is the part that decides whether the quote lands as an investment or as a price.

Not every sale is this sale

The graph above describes a consultative service sale: scheduled labor, a diagnosis, a quote, a signature. Plenty of businesses do not sell that way, and forcing this shape onto them is how you end up with a discovery call nobody wants to be on.

One question sorts it. How much does this buyer need to be diagnosed before they can buy?

The axis that decides the shape of everything after the click

Transactional

They already know what they want. You are competing on trust, availability and friction.

  • Oil change, drain clear, lawn cut
  • A product with a price on the page
  • A $39 template or a monthly plan

The work: remove steps. Every extra field is a lost sale.

Consultative

They know they have a problem. They do not know what the solution is, or what it should cost.

  • The roof that might be flashing
  • A landscape design, a remodel
  • Anything custom, anything over a few thousand dollars

The work: ask better questions than the other two quotes did.

Most businesses are not purely one or the other. A plumber is transactional on a clogged drain and consultative on a repipe, and those two jobs deserve two different paths from the same website. That is a feature, not a contradiction. Sort your services onto this axis before you build a single funnel.

Then there is what you are actually selling

Type one

Services

Done-for-you work on a schedule. Capacity constrained: every job costs you a crew and a day, so a bad fit is expensive twice.

Qualifying is not rude here, it is margin. The referral loop is strongest of the three because the work is visible and local.

Watch: close rate, average ticket, jobs per crew week

Type two

Physical products

Inventory constrained. There is no discovery call. The product page does the presenting and the checkout does the closing.

Conversion levers are photos, reviews, shipping clarity and returns. The second purchase is worth more than the first and almost nobody engineers it.

Watch: conversion rate, cart abandonment, repeat purchase rate

Type three

Digital products and subscriptions

No capacity limit and near-zero cost to serve one more customer. The entire sales process moves inside the product.

Nobody closes anyone. The trial closes them, or it does not. Churn replaces close rate as the number that decides whether you have a business.

Watch: activation rate, time to first value, churn

When nobody picks up the phone

Here is the part the 2021 graph could not show. In a digital or product sale the human interaction with the cycle is minimal or absent, but every stage still happens. Somebody still has to be connected with, diagnosed, shown the fix, and asked for the money. The jobs do not disappear. They get assigned to assets instead of people.

Stage in the service cycleWho does the job when there is no salespersonWhat broken looks like
ConnectThe headline and the first screen. You have a few seconds to say which problem this is for.Bounce rate on the landing page
DiscoverySelf-selection. Pricing tiers, a "which one am I" quiz, use-case pages, an onboarding question or two.People buying the wrong tier, then leaving
PresentThe demo video, the free trial, the interactive tour, the screenshots that show the actual thing.Sign-ups who never open it
CloseCheckout. The guarantee, the card form, the number of fields.Cart abandonment
OnboardNo equivalent in the service cycle. The first session has to deliver something real, or the sale reverses itself quietly.Churn in month one
ReferralShare mechanics and a review prompt fired at the moment of value, not at the end of the month.Silence, and paid traffic forever

So the digital picture is not a loop with a human in the middle. It is a line with a tail, and the tail is where the money is.

Digital and product: the line, not the loop

AttractSearch, social, ads, word of mouth
Offer pageConnect, discover and present, all in one scroll
CheckoutThe close, with no human in it
OnboardFirst value fast, or the sale unwinds
Repeat and referWhere the profit actually lives
Same loop as the service version, different mechanism: the product earns the referral instead of the crew.

The trap in digital

Because there is no salesperson, it is easy to believe there is no sales process, so all the effort goes to the top of the funnel. Then traffic goes up and revenue does not. The stages did not vanish. They are just sitting unbuilt inside your product.

Most of your audience is not in the funnel today

One more thing the graph hides. It implies everyone worth marketing to is somewhere inside it. They are not.

Professor John Dawes at the Ehrenberg-Bass Institute, working with the LinkedIn B2B Institute, put numbers to this. If customers change providers roughly every five years, then about 20 percent are in the market in any given year and only about 5 percent in a given quarter. The other 95 percent are not ignoring your offer. They have no offer to consider yet. (Marketing Week, reporting the Ehrenberg-Bass and LinkedIn B2B Institute research)

That research is business to business, and a homeowner is not a procurement committee. But the mechanic is the same and the cycle is often longer: nobody needs a new roof twice. The practical consequence for a service company is simple. Content that only speaks to people who are ready to buy today is talking to the smallest slice of the room, and the job of everything else is to be remembered in eighteen months when the water heater goes.

Brainstorm your own version with AI

Reading a diagram does not change anything. Saying your own version out loud does, and an AI assistant is a patient thing to say it to. Below are two prompts. Both are built to interview you first rather than hand you a generic answer, because the generic answer is what you already have.

Use the first one if you sell consultative work: quotes, site visits, custom jobs, anything where a human talks to the customer before they buy. Use the second one if you sell a product, a digital product or a subscription and nobody gets on a call.

Prompt one: the four conversations that win service work

This one works through your connect call, your discovery, your presentation and your close, one at a time, then writes you the questions and the language to use.

Copy this prompt
You are going to help me sharpen the four conversations that decide whether my service business wins a job: the connect call, discovery, the presentation, and the close.

Here is my business:
[WHAT I DO, WHO FOR, AND WHERE. ONE OR TWO SENTENCES.]
Typical job size: [DOLLAR RANGE]
How leads reach me: [PHONE, WEB FORM, REFERRAL, OTHER]
When I lose a job, I usually lose it to: [PRICE, A SLOWER RESPONSE, NOT SURE]

Do not give me advice yet.

Interview me first. Ask me ONE question at a time and wait for my answer before asking the next one. Work through the four stages in order, three to five questions each.

1. CONNECT. How I reach a new lead, how long it takes, what I say first, and what I already know about them by the time we speak.
2. DISCOVERY. The questions I ask now, what I tend to find out too late, and what a customer says right before a job goes badly.
3. PRESENTATION. How I deliver a price, what the customer is actually looking at, and what they say back to me most often.
4. CLOSE. What a yes physically looks like, how long it takes to get one, and where deals stall.

If an answer is vague, push once for a specific example from a real job before moving on.

When the interview is done, give me:
- A connect-call opener and three qualifying questions that fit the way my leads actually arrive
- A discovery question bank of eight to twelve questions in the order I should ask them, with the two that matter most flagged and a sentence on why
- The three things my presentation has to say back to the customer in their own words, and what to leave out
- Two closes that suit my job size, plus what to say to the objection I told you I hear most
- The single change that would make the biggest difference, and how I would know within thirty days whether it worked

Keep everything in plain language I would actually say out loud on a job site. Nothing that sounds like a call center script.

Any of those buttons copies the prompt as it opens the tool in a new tab. Paste it in, fill in the brackets, and answer honestly. Vague answers produce vague output.

Prompt two: the path when there is no sales call

For products, digital products and subscriptions. Same interview approach, but it works through the offer page, the checkout, the first hour after purchase and what brings someone back.

Copy this prompt
You are going to help me design the path a stranger takes to becoming a paying customer of something I sell without ever getting on a sales call.

Here is what I sell:
[A PRODUCT, A DIGITAL PRODUCT, OR A SUBSCRIPTION. ONE OR TWO SENTENCES.]
Price: [PRICE OR RANGE]
Who it is for: [WHO THEY ARE AND WHAT THEY ARE TRYING TO GET DONE]
How people find it today: [SEARCH, SOCIAL, ADS, EMAIL LIST, NOWHERE YET]
What happens in the first hour after someone buys: [DESCRIBE IT, EVEN IF THE ANSWER IS NOTHING]

Do not give me advice yet.

Interview me first. Ask me ONE question at a time and wait for my answer before asking the next one. Work through these in order, about three questions each.

1. THE BUYER'S MOMENT. What is happening in their day right before they go looking for something like this.
2. THE OFFER PAGE. What it says now, what proof is on it, and what I know people misunderstand.
3. THE CHECKOUT. Every single step between deciding and paying, counted honestly.
4. THE FIRST HOUR AFTER PURCHASE. What they receive, what they have to do themselves, and what "this worked" looks like from their side.
5. WHAT HAPPENS NEXT. Whether anything brings them back, and whether anyone has ever actually been asked to refer me.

Remember that nobody is going to talk to a salesperson here, so every job a salesperson would do has to be done by the page, the product or the follow-up. Point out any job that currently has nobody and nothing assigned to it.

When the interview is done, give me:
- The one sentence that belongs at the top of the offer page, and why it beats what is there now
- The order the page should make its argument in, section by section, with the proof each section needs
- Every point of friction between wanting it and owning it, ranked by how much it is probably costing me
- A first-session plan: the one thing a new customer must accomplish, and how I get them there
- A repeat-and-refer plan, including the exact moment to ask
- The three numbers I should watch, and what a bad one would be telling me

Be concrete. If you have to assume something about my business, say the assumption out loud instead of guessing silently.

Sell both? Run them both, separately. A repipe and a $39 maintenance plan are two different businesses wearing one logo.

Getting more out of either one

When the AI hands you its answer, reply with: "Now argue the opposite. Where is this advice wrong for a business my size?" The second answer is usually the more useful one.

Score your own funnel in two minutes

Check what you genuinely have today, not what you intend to build. The tally at the bottom updates as you go.

What is actually in place right now

Marketing
The handoff
Sales
The loop

Nothing checked yet. Work down the list.

What to fix first

In order, because the order matters. Each one only pays off if the one above it is already true.

  1. Close the handoff. Measure your real response time for a week before you change anything. Most owners guess an hour and find out it is a day and a half.
  2. Write the discovery questions down. Eight questions on one page, asked the same way every time, is the cheapest close-rate improvement available to a service business.
  3. Fix the Consider stage. Reviews that name the job, photos of work like theirs, a page per service. This moves the same number as doubling ad spend and costs nothing.
  4. Build the referral ask into the job. A fixed moment, a fixed sentence, on every completed job. Not a campaign.
  5. Then buy more traffic. Pouring leads into a funnel with a leak at the seam is how marketing gets blamed for a sales problem.

Questions owners ask

Do I need a CRM for this?

Not at the start. A pipeline is a list of open opportunities with a next action and a date. That can be a spreadsheet. You need a CRM when you cannot remember who is waiting on you, or when more than one person is answering leads. Buying software earlier than that usually produces an expensive, empty database.

My leads all come from referrals. Do I still need the funnel?

You already have one, it just runs on other people. The risk is that it is not yours: it moves with your reputation and stalls when a big referrer retires or moves away. Referral-led businesses are the best candidates for the Consider stage, because a referred prospect almost always looks you up before calling, and what they find decides whether the referral survives the trip.

Where does advertising fit?

Attract only. Ads buy visits. They cannot make anyone believe you, and they cannot answer the phone. If your Consider stage is thin or your response time is slow, ads make the leak bigger and more expensive, which is exactly why so many service businesses conclude that advertising does not work for them.

I sell both a service and a digital product. Which model do I use?

Both, separately. Map each offer onto the consultative-versus-transactional axis first, then give each one its own path from the same site. The common mistake is running the whole business on whichever model your biggest sale uses, which buries the fast, self-serve offer behind a contact form.

How do I know the sales process is the problem and not the marketing?

Compare the number of leads with the number of quotes given, and the number of quotes with the number of jobs won. If leads are healthy and quotes are few, it is a connect or qualifying problem. If quotes are many and wins are few, it is discovery or price positioning. If leads themselves are scarce, then and only then is it a marketing problem.

Sources

Want a second pair of eyes on your funnel?

We help local service businesses find the leak before they spend another dollar at the top. Bring your last twenty leads and we will walk the whole line with you.