How to Do Marketing Attribution: A Step-by-Step Workflow
Doing marketing attribution well is a sequence: get your tracking foundation right, choose a model that fits your business, connect the data sources, and build a reporting habit around it. Skip the foundation and even the fanciest attribution software will produce numbers nobody trusts.
- Step 1: Audit What You're Currently Tracking
- Step 2: Standardize UTM Tagging
- Step 3: Choose Your Attribution Model
- Step 4: Connect Your Data Sources
- Step 5: Build a Reporting Cadence
- Step 6: Layer in Incrementality Testing
- Step 7: Document Assumptions and Revisit Them
- Common Timeline for Getting This Right
This is the workflow we walk clients through at Salterra Digital Services, refined over more than a decade of watching attribution projects succeed or quietly get abandoned six months in because nobody set them up correctly the first time.
Step 1: Audit What You're Currently Tracking
Before adding anything new, find out what’s already broken. Pull up your analytics platform and your CRM side by side and check whether conversions are actually being recorded consistently. It’s common to find double-counted conversions, missing conversion events on key pages like a booked call or a completed checkout, or a CRM where “lead source” is filled in inconsistently by sales reps typing free text instead of selecting from a dropdown.
Make a list of every marketing channel you run — paid search, paid social, organic search, email, affiliate, referral, direct mail, events — and confirm each one has a way to be identified in your data. If a channel can’t be identified, it can’t be attributed, full stop.
Step 2: Standardize UTM Tagging
UTM parameters are still the backbone of digital attribution, and they fail more often from inconsistency than from any technical limitation. Set a naming convention — lowercase, consistent source/medium/campaign structure — and document it somewhere every person who launches a campaign can find it. A campaign tagged Facebook in one place and facebook in another will show up as two separate rows in your reporting and quietly split the credit that channel deserves.
- Use a shared spreadsheet or a UTM builder tool so tagging is generated the same way every time, not typed freehand by whoever is in a hurry that day.
- Tag everything that leaves your control: email campaigns, social posts, influencer links, QR codes, even links in PDFs.
- Never tag internal links between your own pages — this corrupts session data by making it look like a new visit started.
Step 3: Choose Your Attribution Model
With clean data flowing, pick a model that matches your sales cycle and business type. A business with a short cycle and one dominant channel can often get by with last-touch. A B2B business with a six-month sales cycle and five active channels needs a multi-touch model — position-based or data-driven — or last-touch will badly misrepresent which channels are actually building the pipeline.
Don’t treat this as a permanent decision. We typically recommend running two models side by side for the first quarter — usually last-touch and a multi-touch model like linear or position-based — so you can see how differently they characterize the same journeys before committing budget decisions to one view.
Step 4: Connect Your Data Sources
Attribution only works when the data from your ad platforms, your website analytics, and your CRM or sales system can be joined together, usually through a consistent identifier like an email address, a client ID, or a CRM contact record. This typically means:
- Connecting ad platform conversion APIs (Meta Conversions API, Google Enhanced Conversions) so server-side data supplements browser-based tracking that ad blockers and cookie restrictions increasingly degrade.
- Passing UTM and click ID data into your CRM as hidden form fields, so the lead source that shows up at the top of the funnel is still visible when a deal closes months later.
- Using a customer data platform or a tool like HubSpot, Salesforce with Google Analytics integration, or a dedicated attribution platform to stitch these sources into one view.
This is the step most businesses skip, and it’s the reason so many attribution dashboards look impressive but disagree wildly with what sales actually reports closing.
Step 5: Build a Reporting Cadence
Attribution data is only useful if someone looks at it on a schedule and acts on it. Set a monthly review where you compare model output against actual revenue results, and a quarterly deeper review where you check whether the story the data tells still matches what the sales team is hearing anecdotally from prospects — “how did you hear about us” survey responses are a surprisingly good sanity check against your tracked data.
Watch for channels whose attributed value trends in one direction while raw traffic or engagement trends in another. That divergence is often the first sign of a tracking break, not a real performance shift.
Step 6: Layer in Incrementality Testing
Attribution tells you correlation, not causation. To know whether a channel is actually driving incremental revenue — rather than just claiming credit for sales that would have happened anyway — run holdout tests: pause a channel in a specific region or for a specific audience segment and measure whether conversions actually drop. Geo-holdout tests and matched-market tests are the gold standard here, and they’re increasingly necessary as privacy restrictions make pure attribution modeling less reliable on its own.
You don’t need to test every channel constantly. Prioritize testing your largest spend line items first, since that’s where a wrong attribution-driven decision costs the most.
Step 7: Document Assumptions and Revisit Them
Every attribution setup rests on assumptions — the model you chose, the attribution window length, which channels get tracked and which don’t. Write these down somewhere accessible, because six months from now someone will ask why a number looks the way it does, and “that’s just what the model says” is not a satisfying answer to a client or a CFO.
Revisit these assumptions whenever you add a new channel, launch a new product line with a different buying pattern, or notice the model output stops matching lived experience. Attribution setups that are never revisited tend to quietly become less accurate every year as buying behavior shifts and old assumptions stop holding.
Common Timeline for Getting This Right
For most small and midsize businesses, expect the audit and UTM cleanup to take two to four weeks, model selection and data connection another month, and a full quarter before the reporting cadence and incrementality testing produce genuinely trustworthy insight. Attribution is not a project you finish — it’s an operating habit you build and maintain.
Frequently Asked Questions
What's the first thing I should do to set up marketing attribution?
Audit your existing tracking before adding new tools. Confirm every marketing channel is identifiable in your analytics and CRM data, and fix any conversion tracking gaps before layering an attribution model on top of broken data.
How long does it take to get accurate attribution data?
Expect a full quarter before the data is reliable enough to make confident budget decisions. The first month is typically spent on tagging cleanup and tool setup, with the following months building enough conversion volume to trust the model's output.
Do I need special software to do marketing attribution?
No. Google Analytics 4 includes multi-touch attribution reporting for free, and a well-configured CRM can capture lead source data without any additional cost. Dedicated attribution platforms add value once your channel mix and budget justify the expense.
What's the biggest mistake businesses make when setting up attribution?
Connecting an attribution tool before fixing inconsistent UTM tagging and conversion tracking. Clean data discipline matters more than which model or platform you choose.
Should I attribute offline sales too?
If a meaningful share of your revenue closes offline — over the phone, in person, or through a sales rep — yes. Capture lead source at the point of first contact and pass it into your CRM so offline conversions can be tied back to the marketing touchpoint that generated them.
Terry has 30+ years in software and SEO. He’s the founder of Salterra Digital Services and SEO Spring Training, host of the Roundtable SEO Mastermind, and lead instructor at SEO University — teaching the exact tactics his team uses on client work.
Continue learning
Ready to master this?
This guide is one lesson from the Analytics, Measurement & Attribution course. Get every lesson, framework and checklist — plus the full 38-course catalog — inside SEO University.
Practitioner-focused training across the full digital marketing stack — from technical SEO to conversion optimization and the AI search era. By Salterra Digital Services, since 2011.