Meridian, Analytics 360 and QFCs: What the Numbers Actually Mean
Understand the difference between attributed results, modeled contribution and forecasts before the next budget decision.

Meridian estimates how marketing contributes to business outcomes. Analytics 360 brings a Meridian-based budgeting model to selected alpha users. Qualified Future Conversions, or QFCs, estimate later outcomes from early signs of interest. They are related ideas, not three interchangeable ways of counting sales.
The practical question is which decision you are making: explaining yesterday's results, estimating advertising's contribution, or choosing where to invest next. A useful report makes that distinction clear before it recommends a budget.
What changed after Google's announcement
Google's May 2026 announcement connected its marketing mix model with Analytics 360 and described a later connection to QFC signals. The Analytics Help page distinguishes a DDA-based budgeting beta from a Meridian-based alpha for selected Analytics 360 customers. Owning a 360 property does not establish access to the alpha.
The open-source Meridian library is separate. You can use it without buying Analytics 360, but a library is not a finished measurement service: someone still has to prepare the data, specify the model and assess its results.
That difference matters when someone forwards a launch screenshot and asks why the same feature is not in your account. Confirm which product they mean before planning an implementation.
Three questions, three kinds of answer
These views can disagree without one being broken. They may describe different periods, populations or questions. Reconcile those definitions before trying to force the numbers into one total.
For example, a person might discover a retailer through video and buy after a later brand search. A report of the final search interaction, a forecast made earlier in that journey and a model of video's contribution are not three separate customers. Adding them together would manufacture growth.
| Question | Useful evidence | What it does not settle |
|---|---|---|
| Which recorded conversions received advertising credit? | Your conversion and attribution reports. | How many would have happened without the advertising. |
| What did the marketing investment contribute? | An appropriate incrementality experiment or a well-specified causal model. | A certain answer independent of design, data or assumptions. |
| What might happen if we invest differently? | A forecast or budget scenario. | The revenue you have already earned. |
A budget model is useful when you can question it
Meridian works with a business outcome, media activity and relevant non-media factors. Its outputs can include channel contribution estimates, response curves and budget scenarios. Those estimates depend on the model's causal assumptions and on whether the supplied data represents the business well.
Suppose an outdoor retailer increases advertising as a cold spell drives demand for coats. An analysis that cannot distinguish the weather's effect from the advertising may credit the campaign for sales it did not create. This is an illustrative modeling problem, not a claim about a particular Meridian result.
Ask the analyst what changed outside advertising, how those influences were handled, and how sensitive the recommendation is to reasonable alternative assumptions. A narrow-looking chart is not a substitute for that explanation.
Google's model-health guidance also distinguishes fitting the historical data from supporting a causal conclusion. It calls for checks including convergence, negative baseline checks, goodness of fit and how the ROI estimates compare with prior assumptions. A serious failure deserves investigation; a review flag may need business context rather than automatic rejection.
Qualified Future Conversions are not booked orders
Google describes QFCs as estimates informed by ad exposure followed by a leading action, such as a branded search or an engaged visit. The aim is earlier visibility into the long-term effectiveness of upper-funnel media. Google's product explanation describes an out-of-the-box metric rather than a custom-tagging project.
Read that as a prediction, not a receipt. A shopper showing interest has not necessarily paid, received the product or kept it. Do not relabel expected conversions as confirmed purchases in a revenue update.
The May Analytics announcement says these signals will eventually feed Meridian. It does not establish that every current model already uses them. Nor does a public interface illustration establish which campaigns, markets or accounts have access today.

Use the next budget decision to focus the work
Imagine you have another $20,000 to invest. The search report looks strong, the video forecast is promising, and the finance team wants a decision this week. The useful response is not a larger spreadsheet of incomparable metrics.
First define the outcome: incremental revenue, new customers, or contribution after costs. Then ask what the proposed extra spend is expected to achieve, over what period, and with what uncertainty. Check that the comparison includes the costs and channels needed to answer that question.
Choose a manageable change whose effect you can evaluate, rather than funding every channel with an encouraging signal. If the evidence is too weak to separate two options, say so and decide what additional observation or experiment would change the choice.
Fix relevant measurement faults, but do not turn this into a requirement to install every Google data product. A retailer using sales totals for MMM and a lead-generation account importing qualified opportunities have different data needs. The Data Manager guide covers those connection choices separately.
What Parallel can contribute
Parallel can inspect Google Ads conversion-action definitions, their reported campaign-level volume and value, and conversion-lag buckets. That helps you see what the current account reports before comparing it with a separate forecast.
It does not currently have a first-class reader for Meridian models, Analytics 360 budgeting plans or QFC reports. Provide those exports or findings when you want them considered; connecting Google Ads does not silently connect those systems or your CRM.
Use Parallel to examine a concrete discrepancy: which conversion actions explain the reported increase, whether the comparison uses the same period, or which further evidence is missing. A summary should make the difference between observed account data and an external model obvious.
Before you act on the forecast
Analytics budgeting plans are planning tools; creating one does not change budgets in the connected advertising accounts. Its scenario recommendations are tied to the plan's inputs and creation state, so do not reuse an old curve as though it incorporates every later change.
Check the product available to your property, the model behind the result and the date of its inputs. Then translate the recommendation into the actual campaign or business decision it supports. A useful forecast earns a testable decision, not unconditional belief.
Questions and answers
Do I need Analytics 360 to use Meridian?
Not for the open-source library. The integrated Meridian budgeting experience is a separate product with restricted access.
Can I add QFCs to this month's purchases?
No. Do not combine expected future outcomes with recorded sales and present the sum as revenue or orders already earned.
Does a better model fit prove a better budget recommendation?
No. Ask whether the assumptions and business explanation are credible, and whether the model passes the relevant health checks.