Mobile Attribution Explained: How to Know Which Ads Actually Bring You Customers
TLDR: Mobile attribution connects a mobile ad click to the customer action it caused. For businesses without an app, that means tracking calls, forms, and booked jobs. It moves through five links: click, carry, capture, connect, close. Break one link, and the lead lands in your CRM labelled Direct.
A roofing contractor I work with called me from his truck last spring. He had Meta Ads Manager open on his phone showing 47 leads for the month. His CRM showed 19. He wanted to know which number was lying.
Neither. Meta counted everyone who tapped the form and started typing. His CRM counted the ones who finished. Both were right about their own question, and neither answered his, which was how many of those 47 taps ended up as a roof his crew built. That is what job attribution is meant to do. On most accounts I open, nobody finished wiring it.
This is a guide to attribution for people who sell services. Not app installs. Almost everything written about mobile attribution assumes you have something in the App Store, which is useless if what you sell gets booked over the phone.
| Key Takeaways• Mobile attribution is the practice of connecting a mobile ad interaction to the customer action that follows it.• The term usually refers to app installs, but the same logic applies to phone calls, form fills, and booked appointments.• Attribution moves through five stages: click, carry, capture, connect, and close. A failure at any stage loses the record.• Most attribution problems are not analytics problems. They are missing form fields and untracked phone calls.• No setup reaches one hundred percent accuracy. A well-built system typically identifies the source of eighty to ninety percent of leads. |
What Is Mobile Attribution?
Mobile attribution is the practice of tracing a customer back to the ad that produced them. Someone taps something on a phone. Weeks later they pay you. Attribution is the paperwork that says those two people were the same person, and that this ad, not that one, is the reason.
Ask an app marketer and you get a narrower answer. To them it means installs and in-app purchases, tracked through device identifiers and an SDK sitting inside the product. That is the version in the AppsFlyer mobile marketing glossary, and for anyone actually shipping software it is the right one.
Now picture a law firm. A dental practice. A guy with three trucks. No app, no SDK, nowhere to put one. The tracking problem does not go away. It gets worse, because half of what you sell closes on a phone call.
Mobile Attribution vs App Attribution: Which One Do You Need?
Same goal, two entirely different machines. App attribution runs on an SDK buried in your product, device-level identifiers like IDFA on iOS and GAID on Android, and postbacks arriving from a measurement partner. Read that sentence again. If nothing in it sounds like your business, stop worrying about it.
Yours runs on humbler parts. UTM parameters. Click IDs like GCLID and FBCLID. Hidden fields on your forms, a tracking number that swaps by source, and a CRM disciplined enough to hold onto the answer. Adjust publishes a solid technical breakdown of the app side, and it is worth reading if you want the two side by side.
The practical rule is simple. If your conversion is an install, you need a measurement partner. If your conversion is a call, a form, or a booked appointment, you need your CRM wired correctly. Most businesses spending money on Meta and Google fall into the second group and get sold tools built for the first.
How Does Mobile Attribution Work? The Attribution Chain
We describe attribution as a chain of five links. Every link has to hold. If one breaks, the whole record is lost, and the lead lands in your CRM labelled Directly with no source at all.
Link 1: Click
Someone taps your ad. The platform generates a unique click identifier and appends it to your landing page URL along with your UTM parameters. This link almost never fails, because the ad platform handles it.
Link 2: Carry
That identifier has to survive the trip to your website. This is where things start going wrong. Social apps open links inside their own in-app browsers, and referrer data frequently gets stripped in transit. Paid traffic arrives looking like nobody sent it.
Link 3: Capture
When the visitor fills out a form or calls you, the source has to be written down. Hidden fields capture the UTM values. Dynamic number insertion swaps your phone number by traffic source so calls carry a source too. Skip this and every ad dollar becomes untraceable at the exact moment it converts. It is the most common failure we find.
Link 4: Connect
The source has to land on the contact record and stay attached as the lead moves through your pipeline. A source field that gets overwritten during a stage change is the same as no source field at all.
Link 5: Close
When the deal is won, the revenue gets written back to the campaign that started it. Almost nobody does this, and it is the most valuable link in the chain. It is the difference between optimising for cheap leads and optimising for profitable customers, and those are rarely the same campaigns.
Attribution Models Compared
An attribution model is the rule that decides how credit gets divided when a customer touched several ads before converting. Choosing one changes which campaigns look successful, so it matters more than most people assume.
| Attribution model | How credit is assigned | Best suited for | Service business verdict |
| Last-click | All credit to the final touch before the lead comes in | Short buying cycles, one channel | Fine to start with. Undervalues awareness ads badly. |
| First-touch | All credit to the first recorded interaction | Brand awareness and top of funnel testing | Useful for local businesses testing a new channel. |
| Linear | Credit split evenly across every touch | Long, multi-channel journeys | Honest, but tells you very little about what to cut. |
| Time-decay | Recent touches get more credit than older ones | Sales cycles of two weeks or more | Strong fit for home services, legal, and medical. |
| Position-based | Forty percent first, forty percent last, twenty percent middle | Journeys with clear entry and closing points | Best all-round choice for most lead generation. |
| Data-driven | An algorithm weighs each touch using your own data | High volume accounts with steady conversions | Needs real volume. Most small accounts do not qualify. |
Your model choice sits alongside your lookback window, which controls how far back a touchpoint stays eligible for credit. Google documents these settings clearly in the Google Analytics attribution settings guide. Set the window to match your real sales cycle, not the platform default.
How to Set Up Mobile Attribution for a Service Business
You do not need enterprise software for this. You need five things done properly, in this order.
1. Standardise your UTM parameters. Pick one naming convention and never deviate. Lowercase everything. Use facebook, not Facebook and FB and facebook.com, because your CRM will treat those as three separate channels.
2. Add hidden UTM fields to every form on your site. Source, medium, campaign, and content at minimum. This one change fixes more attribution problems than any tool purchase.
3. Turn on dynamic number insertion so calls from ads carry a source. If a meaningful share of your leads phone you, skipping this leaves a hole nothing else can fill.
4. Map those fields to your CRM contact record and lock them so they cannot be overwritten later in the pipeline.
5. Send conversions back to the ad platform’s server-side, and include closed revenue rather than form submissions. This is the step that changes how the algorithms optimise for you.
If you are running paid social alongside this, the tracking and the campaigns need to be built together rather than bolted on afterwards. Our social media marketing packages are set up that way for exactly this reason.
Is Your Attribution Broken? A Seven Point Self-Audit
Answer yes or no to each of these. Be honest, because the point is to find the gaps.
Every form on your site captures utm_source, utm_medium, and utm_campaign in hidden fields.
Your phone number changes based on where the visitor came from.
More than eighty percent of your CRM records show a real lead source rather than Direct or blank.
Closed revenue is written back to the campaign that produced it.
You can name your highest revenue campaign from the last ninety days without opening a dashboard.
Your ad platforms receive conversion events server-side, not only through a browser pixel.
Your attribution window matches how long your customers actually take to decide.
If you answered yes to fewer than five, you are allocating your ad budget on incomplete information. In the accounts we audit, most businesses check two.
Frequently Asked Questions
Why do my leads show as direct traffic?
Direct usually means the referrer data was lost, not that the person typed in your web address. In-app browsers, messaging apps, redirect domains, and missing UTM parameters all produce Direct. A high Direct percentage is a tracking symptom, not a customer insight.
Why does Facebook report more leads than my CRM?
Ad platforms count conversions they can see, using their own attribution window. Your CRM counts records that actually arrived. Some difference is normal. A gap larger than about twenty percent means a link in the chain is broken.
Do I need a mobile measurement partner if I do not have an app?
No. Measurement partners exist to attribute app installs. If your conversions are calls, forms, and appointments, a correctly configured CRM does the same job at a fraction of the cost.
How accurate can mobile attribution actually get?
We have never taken an account past roughly ninety percent of leads with an identified source, and eighty percent is a healthy result. Anyone promising complete accuracy is counting something they should not be. Privacy rules, cross-device journeys, and word of mouth guarantee a permanent blind spot.
How long does it take to set up?
For a single location business with one website and one CRM, most of the work fits inside two weeks. UTM standardisation and hidden form fields can be done in an afternoon. Server-side conversion tracking takes longer and is worth the wait.
Where to Start with Mobile Attribution
The contractor in the truck now has a source on ninety-one percent of his leads. It turned out his cheapest lead source was also his least profitable, a pattern we see constantly and one nobody discovers until the Close link is built. He moved budget, and his cost per signed job fell by roughly a third the following quarter. Nothing about the fix was clever. It was a form with hidden fields, a tracking number, and a CRM configured to hold onto what it already knew.
Run the seven-point audit above. If you checked fewer than five boxes, we will map your attribution chain and show you exactly where it breaks on a free thirty-minute call. You leave with the plan whether you work with us or not. Book a strategy call or send us your questions. We reply in under thirty minutes.
