A Greenwolf Tech Labs site · Our Data Studio agency team, India and USAgreenwolftechlabs.com ↗

Analytics guide

Marketing attribution dashboard: what to build now GA4 has two models

A marketing attribution dashboard shows how much credit each channel deserves for leads, deals and revenue, and how that credit changes under different rules. This guide is for marketing leaders, growth teams and agencies who are tired of every platform claiming the same sale. It covers which attribution models GA4 still offers, how to rebuild the others from raw data, how to join CRM revenue, and how our team builds the dashboard in Data Studio and BigQuery.

Marketer comparing channel credit under different models in a Data Studio marketing attribution dashboard
Analytics guide6 min readUpdated November 18, 2026 · by Harsh Khubchandani

Quick answer

What is a marketing attribution dashboard?

A marketing attribution dashboard is a report that assigns credit for conversions and revenue to the channels and campaigns that touched each customer, then compares spend with attributed results. A useful one reads revenue from the CRM or store, shows at least two models side by side, and explains why channel totals differ from what ad platforms report.

  • GA4 has offered only data-driven and last click models since November 2023.
  • Multi-touch views now need the GA4 BigQuery export or your own touchpoint table.
  • Revenue should come from the CRM or store, not from platform-reported conversions.
  • Show two models side by side and act on channels that rank well under both.
01

What should a marketing attribution dashboard show?

A marketing attribution dashboard should show spend, attributed conversions, attributed revenue and return for each channel, under more than one attribution rule. The comparison is the point: a channel that looks strong under every rule is safe to fund, and one that only looks good under a single rule needs a closer look.

Most teams already have the parts. Ad platforms report spend and their own conversions. GA4 records sessions and key events with a source for each. The CRM or ecommerce platform holds the leads, opportunities, orders and revenue that matter. What is missing is a single table that says, for each conversion, which touches came before it and how credit was split.

We usually build four pages. An overview with spend, revenue and return by channel under the chosen default model. A model comparison page that puts first touch, last touch and a multi-touch rule side by side. A path page with the most common sequences of channels before a conversion and the median days to convert. A campaign page for the people who manage budgets day to day.

  • Overview: spend, attributed revenue, return on ad spend and cost per acquisition by channel.
  • Model comparison: the same conversions credited under two or three rules, with the difference shown.
  • Paths: top channel sequences, touches per conversion and days from first touch to conversion.
  • Campaigns: attributed results by campaign, with platform-reported figures beside them for reference.
02

Which attribution models does GA4 still offer?

GA4 now offers three reporting attribution options: data-driven attribution, paid and organic last click, and Google paid channels last click. Google's Analytics help states that the first click, linear, time decay and position-based models are no longer available as of November 2023.

Google said at the time that fewer than 3% of conversion actions in Google Ads used those four rules-based models, as reported by Ruler Analytics, and made data-driven attribution the default in both Google Ads and GA4. Many attribution dashboard guides still list all six models as menu options, which is no longer true for GA4 reports.

Lookback windows matter as much as the model. According to Google's help on attribution settings, acquisition key events such as first_visit and first_open use a 30-day window by default, which can be cut to 7 days. All other key events default to 90 days, with 30 or 60 days as alternatives. If your sales cycle is longer than 90 days, GA4 alone will not connect the first touch to the sale.

ModelHow credit is splitAvailable in GA4 reports todayHow to get it in a dashboard
Last clickAll credit to the final non-direct touchYes, paid and organic or Google paid onlyGA4 connector in Data Studio
Data-drivenCredit from Google's model of which touches raised conversion oddsYes, the defaultGA4 connector, as a black box
First touchAll credit to the first known touchNot as a model, but first user source existsGA4 first user dimensions or BigQuery
LinearEqual credit to every touchNo, removed in 2023Rebuild in BigQuery from event data
Position-based40% first, 40% last, 20% shared by the middleNo, removed in 2023Rebuild in BigQuery from event data
Time decayMore credit to touches closer to conversionNo, removed in 2023Rebuild in BigQuery from event data
03

How do you build a multi-touch attribution dashboard without GA4 models?

You build a multi-touch attribution dashboard by exporting event-level data to BigQuery, assembling each converter's touches into a path, and applying your own credit rules in SQL. Data Studio then reads the finished tables.

GA4's BigQuery export is the usual starting point. Google's help says standard properties have a daily export limit of 1 million events, and the streaming export costs $0.05 per gigabyte, roughly 600,000 events. Storage and queries are billed by BigQuery, so we partition tables by date and build daily summary tables rather than letting every chart scan raw events.

From the raw events we build a touchpoint table: one row per user, session and source, with timestamps. Users are stitched with a user ID or a hashed CRM identifier where consent allows it. Conversions come from a second table, ideally the CRM or store rather than GA4, so revenue is the figure finance recognises. A scheduled query then joins the two inside your chosen window and writes credit per touch for each model.

  • Turn on the GA4 BigQuery export and keep the raw dataset untouched.
  • Build a touchpoint table with user key, timestamp, source, medium and campaign.
  • Load conversions and revenue from the CRM or ecommerce platform with the same user key.
  • Write credit per touch for each model in a scheduled query, inside a fixed window.
  • Add spend by channel and campaign from ad platforms, then point Data Studio at the outputs.
04

What does a worked attribution example look like?

A worked example shows why the model choice moves budget. Take an illustrative B2B deal worth $12,000 with four touches over 41 days: a LinkedIn ad, an organic search visit, a webinar email and a branded Google Ads click before the demo request.

Under last click, branded search gets the full $12,000 and the other three channels get nothing. Under first touch, LinkedIn gets the full $12,000. Under linear, each touch gets $3,000. Under position-based, LinkedIn and branded search get $4,800 each, while organic search and email get $1,200 each.

Now add spend. If branded search cost $40 for that click and the LinkedIn campaign cost $2,000 per closed deal on average, last click makes LinkedIn look like pure cost and branded search look like a 300 to 1 return. Position-based gives a far more believable picture. A good attribution dashboard shows both views so the team can see that branded search mostly harvests demand that other channels created.

TouchLast clickFirst touchLinearPosition-based
LinkedIn ad (day 1)$0$12,000$3,000$4,800
Organic search (day 9)$0$0$3,000$1,200
Webinar email (day 30)$0$0$3,000$1,200
Branded Google Ads (day 41)$12,000$0$3,000$4,800
05

Why do ad platforms and attribution dashboards disagree?

Ad platforms and attribution dashboards disagree because each platform credits itself for any conversion it touched, under its own window and counting rules. Add up Google Ads, Meta and LinkedIn conversions and the total will often exceed the number of real customers in the CRM.

View-through conversions widen the gap, since a platform may claim a sale from an ad that was only seen. Different lookback windows and time zones shift conversions between days and months. And platforms report conversion value as the tag recorded it, which may not match the deal value after discounts, refunds or a sales negotiation.

The fix is to treat the CRM or store as the one count of customers and revenue, and to show platform-reported conversions on a separate column labelled as such. That way, managers can still optimise inside each platform while budget decisions use one deduplicated number.

06

How do CRM revenue and offline deals fit into attribution?

CRM revenue fits in by joining each closed deal back to the touches recorded for that contact. Without that join, an attribution dashboard can only credit form fills, which is weak for any business with a sales team.

The join needs a shared key. The cleanest approach is to capture the GA4 client ID or a user ID in a hidden form field and store it on the CRM contact, along with UTM values and the original source. Phone calls need call tracking numbers per channel. Deals that cannot be matched should be reported as unattributed rather than quietly given to direct or to the last channel.

For long sales cycles, report by cohort. Group leads by the month they arrived and show the pipeline and revenue they produced later. A campaign judged only on revenue that closed in the same month will always look worse than one that ran a quarter earlier.

07

Can you build a marketing attribution dashboard in Data Studio?

Yes, you can build a marketing attribution dashboard in Data Studio, but the attribution logic should live upstream. Data Studio is strong at presenting the output and weak at path modelling.

Blends in Data Studio are limited to 5 data sources with equality joins, so joining spend from several ad platforms, GA4 and a CRM inside a blend quickly reaches the ceiling. Calculated fields and CASE statements cannot rebuild a path from event-level data. And many charts reading the GA4 connector at once can hit GA4 API quotas. For a simple first-touch versus last-click view, the GA4 connector's first user source and session source fields are enough. For anything multi-touch, we model in BigQuery and give Data Studio one table per page.

Data Studio then adds what it does well: a model selector built with a parameter, date and channel filters, scheduled email delivery to leadership, and Conversational Analytics, generally available since July 30, 2026, built on top of BigQuery data agents so managers can ask questions in plain language.

08

What are the common mistakes, and when is attribution not worth it?

The most common mistake is choosing a model before fixing tracking. If UTMs are inconsistent, form fills are not stored on the CRM contact, or half the conversions arrive by phone with no tracking, no model will give a trustworthy answer.

Other mistakes we see: treating data-driven attribution as proof rather than an estimate, reporting only one model, mixing platform-reported and CRM conversions in one total, and ignoring channels that never get clicked, such as podcasts or offline events. Attribution measures recorded touches, so it understates anything that builds demand without a click.

Attribution is not worth a custom build if you run one or two channels, if most sales come from a few large deals that sales already understands well, or if monthly conversions are so few that model differences are noise. In those cases, a clean last-click report with CRM revenue is enough. Teams with heavy budgets and long cycles often pair attribution with holdout or geo tests to check the model against reality.

FAQ

Frequently asked questions

What is the best attribution model for a marketing dashboard?

No single model is best. Show last click and a multi-touch rule such as position-based side by side, and fund channels that perform well under both. Data-driven attribution in GA4 is a useful third view, but it cannot be inspected or recalculated outside Google.

Does GA4 still have linear and position-based attribution?

No. Google's Analytics help states that first click, linear, time decay and position-based models are no longer available as of November 2023. GA4 now offers data-driven attribution and two last click options. You can rebuild the other models from the GA4 BigQuery export.

How do I build a multi-touch attribution report in Data Studio?

Export GA4 events to BigQuery, build a touchpoint table per user, join CRM or store conversions, and write credit per touch for each model with scheduled SQL. Then connect Data Studio to the output tables and add a parameter to switch models.

Why don't my Google Ads and Meta conversions match my CRM?

Each platform credits itself for conversions it touched, including view-through conversions, using its own window and time zone. Added together they often exceed real customers. Use the CRM or store as the single count and show platform figures separately.

How long is the GA4 attribution lookback window?

For acquisition key events such as first_visit, the default is 30 days and can be set to 7. For all other key events, the default is 90 days, with 30 or 60 days as options. Longer sales cycles need a CRM or BigQuery model to connect early touches.

Do I need BigQuery for a marketing attribution dashboard?

Not for a simple first-touch and last-click view, which the GA4 connector can provide. For linear, position-based or time decay models, CRM revenue joins and paths, BigQuery is the practical route. Standard GA4 properties can export up to 1 million events a day.

Can an agency show attribution to clients in a white-label report?

Yes. The model tables are built once per client in BigQuery, and a shared Data Studio template with the agency's branding reads each client's tables. Label platform-reported figures clearly so clients do not compare them with attributed totals.

Reviewed by Team Greenwolf Tech Labs · 18 Nov 2026

Get started

Tell us what your reporting has to do

Describe the dashboards you need, who reads them and where the data sits. We reply within one business day with an approach, the connectors involved and a fixed-price plan.

  • Free 30-minute reporting review
  • Fixed quote before any work starts
  • Everything built and owned in your Google account

Prefer email? info@greenwolftechlabs.com

Get a free proposal

Two lines is enough to start.

A Data Studio lead replies from info@greenwolftechlabs.com, usually within one business day. No mailing lists.

Ask us anything

A Data Studio lead replies from info@greenwolftechlabs.com, usually within one business day.

A Data Studio lead replies from info@greenwolftechlabs.com, usually within one business day. No mailing lists.