What Is Attribution Analysis? Making Indirect Impact Visible and Optimising Budget

“Our CPA on online advertising is within target, yet total lead numbers have stalled.”
“The effect of our owned media is unclear, so we have no strong case for the budget.”

If this is your problem, the cause may be that your measures lean too heavily on the last click, the final touchpoint。

Why is judging on the final touchpoint alone not enough?

In a B2B buying process it is very rare for a customer to close on a single visit. From awareness through consideration to a decision, they move back and forth across a great many touchpoints: search, social media, trade shows, email newsletters.

Judging only the touchpoint where the conversion happened is, in football terms, praising the player who scored while ignoring the one who made the decisive pass.

Stop investing in the player making the pass — the awareness and consideration work — and you fall into a vicious circle where the whole team scores less, which is to say you win fewer leads.

What matters here is a way of fairly crediting every touchpoint on the way to a result: attribution analysis.。

This article unpicks the increasingly complex customer journey and explains how to identify the channels and content genuinely worth investing in.
Use it as a practical guideline for maximising marketing ROI across the organisation.

Summary
・Built for the complex paths particular to B2B: essential in a B2B business with long consideration periods and many touchpoints, both to optimise budget allocation and to stop the future customer pool drying up.
・Use the five models for different jobs: you need to pick the right rule for the business phase and strategy — first interaction, which favours awareness, or position-based, the U-shape.
・Doing it in GA4, and what to watch: covers everything from the analysis procedure using GA4’s standard features to the technical background, such as the limits on measurement that recent cookie restrictions (ITP and the like) impose.

▼Who this is for
・Marketing leads whose revenue has stopped growing
・Anyone who wants to put numbers on the indirect contribution of owned media and trade shows
・Anyone who wants to decide from data which channel to concentrate budget on
・Anyone who wants to know how to do attribution analysis in GA4

1. What is attribution analysis?

We start with the basic definition of attribution analysis and why it has become essential in B2B marketing today.

Making each stage’s contribution to the result visible

「Attribution analysis is a method that assigns a share of the contribution to every ad, publication and piece of content the user touched on the way to a conversion, and analyses it。

By defining which activity a result is attributed to, you can properly judge the value of work that has been overlooked until now.

Conventional web analytics generally used the last-click model, which credits only the source immediately before the result. Attribution analysis turns the spotlight on the first touchpoint, where awareness starts, and on the middle ones, where understanding and consideration happen.

For companies working with demand generation — the framework that runs from creating prospects through nurturing leads to passing the ready-to-buy ones to sales — attribution analysis is indispensable for understanding where and which content contributed to a conversion.

Related article:
What is demand generation? Concrete methods and implementation steps for maximising leads, meetings and revenue

1.2 Why attribution analysis is needed now

The reason is that the customer’s buying process has grown more complex.
With B2B products in particular, the following three reasons mean that judging on the last click alone no longer gives you an accurate picture。

① Longer consideration periods and more touchpoints

It is not unusual for a B2B product to be under consideration for several months or more than a year. Over that time the person in charge meets information by many routes.

• Reading a problem-solving article found in a search engine
• Checking your reputation on social media
• Downloading a case study guide through an ad
• Finally searching the company name and making an enquiry

In this sequence, if you credit only the final search on the company name, the value of the article that made first contact and of the guide that built trust drops out of the calculation。

② Keeping the prospect pool from drying up

Make the last click your only KPI and budget tends to concentrate on particular advertising, such as search advertising aimed at people ready to buy now.
That looks efficient in the short term, but awareness work aimed at new prospects gets neglected, so the pool of future candidates dries up。
The result is a rising cost per acquisition (CPA) and growth that stops.

③ Proving the ROI of content

Owned media, video and white papers rarely produce a direct result, but they play an important part in nurturing customers.
If you cannot put numbers on that assist, the content budget becomes a candidate for cuts.
Attribution analysis lets you put numbers on that assist — how much this work pushed an order over the line。

Understanding complex acquisition paths correctly is therefore unavoidable if you want to optimise the marketing budget and maximise revenue.
Aggregating a huge amount of data by hand and assigning a contribution to each touchpoint is very hard, though. What works is to use the analysis features in a CRM or MA tool.

HubSpot, which has a large share worldwide, has features that let you do advanced analysis without complicated set-up. If you want a clearer picture of how the tool is used, see the following article.

Related article:
What is attribution analysis in HubSpot? Understand how much your marketing contributes to revenue

2. Before you start: getting the data in order

If the data you put into attribution analysis is inaccurate, the results you get out will be wrong.
Here we explain the data groundwork a marketing lead must check before beginning.

2.1 Redefining conversion points and setting micro conversions

First, define clearly what counts as a result. If you set only the final goals — an enquiry, or a request for a meeting — as conversions, there may not be enough data for the analysis, which makes a statistical judgement difficult.

To make attribution analysis more accurate, we recommend setting micro conversions — the points in between — as measurement points, like these。

• Downloading a white paper
• Signing up to the email newsletter
• Viewing the pricing page
• Registering for a webinar

Understanding what users do at these intermediate points makes visible which activity is lifting their interest, and how.

2.2 [Important] Managing tracking parameters (UTM parameters) rigorously

The single biggest factor in the accuracy of attribution analysis is identifying the source accurately.

Tools such as Google Analytics 4 (GA4) identify some sources automatically, but email newsletters, QR codes and links inside sales material are treated as unknown data unless you tag them properly.

So it is better to write it as For accurate analysis, make adding the following URL parameters a rule across the organisation and stick to it。

Parameter	What it does	Example value
utm_source	Source (where from)	google／facebook／newsletter／ qr_code
utm_medium	Medium (by what means)	cpc／organic／email／display／ social
utm_campaign	Campaign (under which initiative)	2024summer_sale／ whitepaper_dl

The point to watch most closely here is preventing inconsistent spellings
If “facebook” and “Facebook” both appear, the tool treats them as separate channels and the data cannot be combined。
Create a management sheet and put in place a set-up that runs one rule across the whole company.

2.3 Defining channel groups to suit your business

Tools such as GA4 come with default channel groups, but you may need to customise them for your own business model.

For example, manage branded search — searches on your company or service name — separately from generic search, where people search on a problem keyword.
Branded search is the behaviour of people who already know you; generic search is the behaviour of people looking for a new point of contact.
Lump them together as paid search and you can no longer judge properly how well the advertising actually opens up new ground.

3. The five main attribution models and how to choose

Attribution analysis comes in several patterns, or models.
Choosing the right model for your marketing strategy and the nature of your product matters.
Here we explain the five main models and where each is used.

Model	How it works	Advantages	Disadvantages
① Last interaction	Credits the final touchpoint with 100%	Makes clear the final deciding factor that led straight to the result	Ignores the contribution of the awareness and consideration stages
② First interaction	Credits the first touchpoint with 100%	Properly credits the work that triggered awareness	Hides the contribution of the nurturing and closing that followed
③ Linear	Splits the contribution equally across every touchpoint	Judges the whole process flatly, which prevents anything being undervalued	The important points — the trigger and the decider — tend to get buried
④ Time decay	Credits touchpoints nearer the result more highly and older ones less	Credits the work that gave a push once buying intent had risen	Underestimates the contribution of the awareness that first sparked interest
⑤ Position-based (U-shaped)	Weights the first and last touchpoints heavily and credits the middle as supporting	Best suited to B2B. Weights both ends, awareness and the order, while still giving an overall view	The work in the comparison and consideration phase, in the middle, may be undervalued

3.1 ① The last interaction model

The last interaction model gives 100% of the value to the ad or link clicked immediately before the order。

It suits products with a short consideration period that are bought on impulse.

In a business with a long consideration period, though, take care: every earlier contribution — reading the blog, requesting a document — is ignored.

3.2 ② The first interaction model

The first interaction model gives 100% of the value to the first touchpoint that set the order in motion。

It is useful in a phase where getting known comes first.

It cannot measure, however, how much the later newsletters and follow-up by salespeople — the nurturing — contributed to the order.

3.3 ③ The linear model

The linear model spreads the value equally across every channel touched。

It suits a strategy that keeps in steady contact with customers rather than leaning on any one activity.

The overall picture is easy to grasp, but the drawback is that it is hard to see which activity actually clinched the order.

3.4 ④ The time decay model

The time decay model credits touchpoints nearer the date of the order more highly and lowers the credit the further back you go。

It suits measuring the effect of a short promotion or of the work that gives a push just before a meeting.

Because the awareness work that brought the customer in first is credited less, take care not to misjudge how important new traffic is.

3.5 ⑤ The position-based (U-shaped) model

The position-based (U-shaped) model assigns 40% each to the first and last touchpoints and 20% across the ones in between。

For most B2B companies it is the best-balanced model.

It shows clearly where the lead came from and where it became a meeting, which makes it ideal for an organisation where marketing and sales have to work together.
Content in the middle phase tends to be undervalued, so combining it with the linear model to fill that gap is the professional approach.

4. In practice: the analysis procedure in GA4 and how to read it

With the theory in hand, here is the procedure for actually running the analysis in Google Analytics 4 (GA4).
GA4 includes attribution analysis as standard, in the advertising workspace, so anyone can start doing advanced analysis.

4.1 Step 1: open the Advertising section

① Click [Advertising] in the left-hand menu of GA4

② Click [Attribution models]
※ This is the main screen for attribution analysis
※ Conversions and the links to Google Ads and the like must be set up beforehand

4.2 Step 2: find the hidden contributions with model comparison

GA4’s model comparison feature lets you change the rule — the model — and so draw out how much a particular activity is holding up the result.

① Select [Attribution models]
※ Here you can place different attribution models side by side and compare the numbers

▼Attribution models (the rule for crediting)

① Select [Attribution paths]

▼Attribution paths (the history of behaviour)

The key to the analysis is to look at the difference between the numbers.
Put last click and linear side by side and check how conversions and cost per acquisition (CPA) change.

For example
Say the results for display advertising or organic social are more than 20% higher under the linear model than under last click.

What that tells you: this work rarely becomes the last click, the direct order, but it is working as a powerful assist that widens awareness and deepens consideration.
It gives management something to decide on: cut the budget purely because there are few direct orders and total orders may drop sharply a few months later.

4.3 Step 3: find the winning patterns with conversion paths

Beyond comparing numbers, following the actual footsteps a user left on the way to an order reveals the ideal customer journey.

▼Example actions

・Look at the length of the path
Check how many touches it takes on average to reach a conversion. The more touches, the more sides your approach needs.

・Spot the patterns
If the pattern “organic search (article views) × 2 → paid search (branded)” comes up often, that is evidence your SEO articles are working as nurturing.

・Identify the first touchpoint
Identify the channel that appears most often at the start of the path. That is where your leads come from.

Related article:
What is attribution analysis in HubSpot? Understand how much your marketing contributes to revenue

5. Points to watch: cookie restrictions and the limits of measurement

Attribution analysis is a powerful weapon, but technical constraints are making it harder to capture all the data with complete accuracy.
Understanding these limits properly matters a great deal when making sound management decisions.

5.1 The effect of ITP (Intelligent Tracking Prevention)

Built into browsers such as Apple’s Safari, ITP places tight restrictions on the use of cookies to track what users do。
Third-party cookies in particular are tightly restricted, and even first-party cookies tend to have their lifetime shortened, to a day or seven days, for example.

Say a user reads an article on day one, then a month later searches the company name and makes an enquiry.
Those two would once have been linked as one person’s behaviour, but now that cookie lifetimes are shorter the system may treat them as visits by different people.
As a result, the assist effect over a long consideration process tends to be undervalued in the data。

5.2 The break in cross-device behaviour

Behaviour often seen in B2B — gathering information on a phone during the commute, then making the formal enquiry from a work computer on arrival — is another barrier to measurement.

Unless you can identify the user by a specific ID, such as login details, the browsing history on the phone and the order data on the computer are split apart and counted as two different users。

5.3 How to work with data from here

What matters is not chasing perfect numbers.
Keep the following two points in mind and treat the results as a compass for decisions。

・Weight the broad trend and the relative balance
Rather than taking heart or losing it over the last digit, concentrate on grasping the trend: which channel is growing, and which activities move together.

・Use AI estimation
Technology in which AI infers and fills in missing data, as GA4’s data-driven attribution does, is advancing. Use the latest features properly and accurate forecasting becomes possible.

6. Analysis is a means of making better decisions

The goal of attribution analysis is not to produce a finely detailed report.
On the basis of the results,

“optimise the budget allocation”
“decide the direction of the content to focus on”

and the like —taking concrete management action of this kind is the real purpose

In B2B marketing it is rare for a single activity alone to bring a customer to an order.
They read an article, see an ad, receive an email, attend a webinar.
A meeting comes at the end of that patient accumulation of touchpoints。

Not just the final push immediately before the order, but finding the unsung contributors that carried the customer’s change of mind along the way, and crediting them fairly. That is the key to raising the return on all your marketing and achieving sustained revenue growth.

Start by opening the model comparison report in GA4, placing your current model beside others, and checking which activity’s numbers change most.
In that difference lies the next hint for taking your business further.