Multi touch attribution credits a sale to every recorded interaction that led to it, not just the first or the last. You will also see it written as MTA. For a firm whose clients take weeks to decide, it is the only honest picture of what your marketing does. This article answers three questions in order: whether MTA or media mix modelling suits a business of 10 to 250 people, how to capture touches on the levelpartner.ai platform so every won deal carries a full history, and how to turn that history into a multi touch attribution model your team will run every quarter.
What a touch record looks like
Take a typical consultancy client. They click a LinkedIn ad and download a guide. Two weeks later they open a nurture email and click through to a case study. They attend a webinar. They reply to a follow-up text asking for a call, book a slot through your calendar and sign a month later.
That is six touches across four channels. A single-touch report shows one of them. A multi-touch record shows all six on the contact's activity history, each with a date and a channel. The record is the foundation; the model is only a rule applied to it.
MTA and media mix modelling
There are two serious ways of measuring marketing, and finance directors tend to ask which one the business needs once the acronyms appear.
MTA works from the bottom up. It follows identified people through their recorded touches and splits the credit for each sale among them. It needs person-level data and is only as good as the tracking behind it.
Media mix modelling, or MMM, works from the top down. It takes total spend and total sales per week over two years or more and estimates statistically how much each channel contributed. It needs no person-level tracking, but it needs long history and plenty of variation in spend to say anything useful.
For a firm selling consulting programmes, agency retainers, mortgages or property services, the answer is MTA almost every time. You win tens or hundreds of clients a year, not millions of transactions, so MMM would return wide guesses dressed up as precision. MTA gets more useful the more personal and considered the sale is, because each client already has a traceable history in the CRM.
Suppose a firm wins 140 new clients a year on an illustrative £400,000 marketing budget across six channels. An MMM study would have about 100 weekly data points and small spend changes per channel, too little to separate their effects. A quarterly MTA review would cover around 35 won clients, each with eight to twelve recorded touches. On that comparison, MTA is the clear choice.
What MTA needs to work
Three conditions. Every contact must be identifiable across touches, which means capturing an email or phone number early. Every touch must carry a channel label, which means tracking parameters on links and tracking numbers on calls. And revenue must be recorded against the same contact, which means a pipeline the sales team actually uses. The platform holds all three together: contacts, their activity history and opportunities. That is why MTA is achievable at this size without buying a separate product.
Multi touch marketing attribution versus sales activity
Keep marketing touches and sales activity apart. Marketing touches build interest: ads, content, emails, events. Sales activity happens once a person is talking to a salesperson: calls, meetings, proposals. Multi touch marketing attribution should count the first group, plus the booking that bridges into sales. Counting every sales call hands most of the credit to the sales process and hides the marketing behind it. Credit for salespeople belongs in sales attribution, which reads the same records for a different question.
Before you start
- A written definition of a touch that counts: clicks, form fills, calls, bookings and replies.
- Contact capture early, ideally at the first form or download.
- Tracking parameters on every outbound link and tracking numbers by channel for calls.
- Webinar and event attendance lists you can import against contacts.
- Won values and close dates on every closed opportunity.
How to set it up
- Write down which interactions count as a touch.
- Go to Contacts and select a recent won client's record.
- Read the attribution data and history under the Activity tab.
- Patch any channel with missing touches, starting with phone calls.
- Import each webinar attendee list and tag the imported contacts.
- Choose a weighting model: linear, time decay or position-based.
- At quarter end, list each won client's touches from their Activity tab.
- Split each deal's value across its touches in a spreadsheet.
- Compare the totals with a First Attribution widget on your dashboard.
- Repeat each quarter under the same rules.
The comparison widget in step 9 is an Opportunity widget with an Attribution condition set to First Attribution and grouped by session source. The attribution model guide shows how to build it.
Say first touch credits an illustrative £180,000 of a quarter's won work to paid social and £20,000 to the webinar series. After webinar attendance is imported and a position-based split applied, imagine paid social holding £120,000 and webinars rising to £65,000. A shift like that would argue for keeping a webinar budget that was due to be cut, and reviewing the paid social targeting instead.
Choosing a multi touch attribution model
The three common models each lean a different way. Linear treats every touch equally. Time decay gives more weight to touches near the sale. Position-based gives most credit to the first and last touches and spreads the rest across the middle. For a 30 to 90 day cycle, start with position-based. For a moderate cycle with steady nurture, linear is easier to explain, and linear credit in practice walks through the arithmetic.
What most articles get wrong
They declare MTA dead because third-party cookies are fading. That is true for advertisers following anonymous people across the web. It is far less true for a service firm, because your most important touches happen after someone has given you their details: emails, calls, meetings, texts. Those are first-party records in your own CRM, and no browser change removes them.
Dedicated tools and when they make sense
Specialist products are marketed for this job, such as AppsFlyer multi touch attribution for app marketers, or measurement services from Nielsen and Neustar aimed at large advertisers. Before evaluating any of them, check three things: whether clients convert through an app or through conversations, whether your spend is large enough to need outside measurement, and whether the tool can read the calls, meetings and won deals held in your CRM.
Buying an MTA tool before the CRM records touches properly, or counting email opens as touches. A specialist tool cannot credit calls that were never tracked or deals never marked won. Opens are unreliable on many devices, and a weekly newsletter piles up more opens than any other channel produces touches. Fix capture first and count clicks, replies, form fills, calls and bookings only.
Time it takes
Defining touches and patching tracking takes two to three days of part-time effort. The quarterly calculation takes one to two hours once the rules are set. There is no daily effort.
When not to use it
If most won clients have one or two touches, a multi-touch model reproduces the single-touch view; use it once four or more meaningful touches per client are routine. If most spend goes on brand media with no response mechanism, MTA undercounts it by design, so measure that spend the way media attribution describes.
Troubleshooting
Most histories start at the booking. People browse anonymously until they book. Add an earlier reason to share an email address, such as a short guide.
One channel dominates every model. Reminder or service messages are probably being counted. Exclude them from your touch definition.
Touches sit under the wrong channel. Tracking values are inconsistent. Agree one naming list and correct live campaigns.
The result is hard to explain. The weighting rule is too elaborate. Fall back to linear or position-based, which anyone can follow on paper.
Frequently asked questions
Is MTA the same as multi touch attribution?
Yes. MTA is the abbreviation, and vendor material uses both. The method and the data it needs are identical.
Can MTA and MMM be used together?
Large advertisers do, using MMM for budget levels and MTA for tactics. For a firm of 10 to 250 people, the combination is rarely worth the cost.
How far back should the data go?
At least two full sales cycles. For a 60-day cycle, use a minimum of four months of won deals.
Can the platform show multi-touch results natively?
The platform stores first and latest attribution on each contact, and dashboard widgets can filter or group by either. The full history sits on each contact record, so the multi-touch split itself is calculated outside the platform each quarter.