Marketing attribution: a first touch model for small teams

Marketing attribution is how you decide which channel gets credit for a sale. First touch attribution gives the credit to the channel that first brought a person in, last touch gives it to the last thing they clicked before buying, and multi touch models split it across several steps. For a small team, first touch attribution with revenue taken from real payments is a sound default, because it answers the question that drives your ad budget: where do buyers come from?

One buyer, three answers

Take a consultant who sells a $3,000 program. One buyer's path looks like this:

1. Clicks a Meta ad and downloads a free guide.
2. Gets a few emails and a text conversation over two weeks.
3. Registers for a webinar from one of the emails and attends.
4. A week later, searches the consultant's name on Google, clicks a search ad and books a call.
5. Pays after the call.

Ask three models who gets the credit and you get three answers.

• First touch: the Meta ad gets the full $3,000.
• Last touch: the Google search ad gets the full $3,000.
• Linear multi touch: Meta, email, the webinar and Google each get $750.

Each model answers a different question.

The main attribution models

First touch

All credit goes to the first tracked interaction. It answers: which channels bring in people who end up buying? That is the question behind most budget decisions in a small business, because the ad that first brings someone in is what you pay to scale.

Last touch

All credit goes to the last tracked interaction before the sale. It answers: what was the final step before people bought? In practice, last touch tends to credit branded search, email and retargeting, the channels that collect demand other channels created. In the example above, nobody would have searched the consultant's name without the Meta ad.

Multi touch

Credit is split across touches by a rule. Linear splits it evenly, time decay gives more to recent touches, position based gives most to the first and last, and data driven models let software decide the split from patterns across many buyers.

Multi touch looks more complete, and it works for companies with thousands of conversions and tracking that follows people across devices. With 20 or 50 sales a month, the splits rest on very few data points, and small changes in tracking move credit around more than real changes in performance do.

Why first touch is a sound default for small teams

First touch has real limits, covered below. For a coach, course creator, agency or community owner spending a few thousand dollars a month on ads, it has four things going for it:

• It answers the budget question directly: which channels produce buyers.
• It is easy to audit. One person, one channel, one date. If a number looks wrong, you can open the list of buyers and check them one by one.
• It does not need perfect tracking of every later step, which matters when people move between phone and laptop, email and text.
• It stops harvest channels like branded search and email from looking like the reason people buy.

What first touch misses

First touch gives no credit to the steps that convert people once they arrive: your webinar, your emails, your text follow ups, your sales calls. Those steps matter enormously, and you should measure them directly with funnel numbers: registrations, attendance, calls booked, calls held and purchases. Attribution tells you where to find buyers. Funnel numbers tell you how well you turn people into buyers.

Take revenue from payments, not from forms

An attribution report is only as good as its revenue figure. Use payments from your payment processor, matched to each contact, with refunds subtracted. Do not count a booked call or a signed proposal as revenue until money arrives.

Match payments to people by exact email first, then exact phone number. Never match by name: two people called Sarah Miller are not the same buyer.

Count only payments made on or after the person's first tracked touch. A client who paid you for two years and then clicked an ad would otherwise make that ad look like it produced two years of revenue. Keep two numbers on each contact: lifetime revenue (everything they paid) and attributed revenue (what they paid from their first touch on). Only the second one belongs in attribution.

Untagged traffic

Some visits arrive with no readable source. These are your untagged visits, and every report has some. Common causes:

• An ad link without tracking tags, or with tags the platform was supposed to fill in automatically but did not.
• Links shared in private messages or group chats, which lose their referrer.
• Redirects that strip the tags off the end of the link.
• In app browsers that pass little or no referral information.

You will never get untagged to zero, but you can keep it small:

1. Tag every link you control, in every ad, post, email and bio.
2. Use one spelling per source, always lowercase, and write the convention down.
3. Check a sample of live ad links to make sure dynamic tags are being filled in.
4. Use short links that keep the tags intact when you post in places where long links look messy.
5. Watch the untagged share week to week, and look into any sudden jump.

A simple tagging convention covers most small teams: utm_source is the platform (facebook, instagram, google, tiktok, youtube, newsletter), utm_medium is the type (paid, organic, email, sms), utm_campaign is the campaign name, and utm_content is the ad or post.

Send real sales back to the ad platforms

Ad platforms decide who sees your ads based on the results you report to them. If you only report leads, the platform finds people likely to fill in forms, which is not the same group as people likely to pay $3,000.

Reporting real purchases, with the amount actually paid, for buyers who came from that platform, gives the platform's targeting the signal you care about. Two rules keep this honest:

• Do not report organic sales as paid conversions. A sale from a referral, reported to Meta as a Meta purchase, teaches Meta to look for people who were never going to come from Meta.
• Check amounts before you send them. Treat a reported purchase as permanent, because platforms count what they receive.

A first touch setup checklist

1. Tag every paid link and every link you post yourself.
2. Record each person's source on their first visit, and keep it as their first touch even when later visits come from other channels.
3. Keep the later touches on the record for context.
4. Match payments to contacts by exact email, then exact phone.
5. Count attributed revenue (paid on or after the first touch) after refunds.
6. Report by channel: people, buyers, conversion rate, revenue and revenue per person.
7. Review untagged visits every week.
8. Send qualified purchases back to the platform each buyer came from.

How Niche Engine handles it

Attribution in Niche Engine groups every contact by their first tracked touch and shows people, buyers, conversion rate, attributed revenue and revenue per person for each channel. The header counts people who touched more than one channel, and a tile shows Untagged visits, whose ad tag the platform never filled in. Click a channel to see its top people by attributed revenue, with campaign and first touch date.

Revenue comes from Stripe, matched to each contact by exact email, then exact phone, never by name, with refunds taken off. Attributed revenue counts only payments on or after the first tracked touch, and lifetime revenue is kept separately. Qualified purchases are reported back to Meta, Google, TikTok and OpenAI ads, and organic sales are never reported as paid conversions. In Niche Content, short links keep your tracking tags tidy so visits are measured where they land.