← Back to Knowledge Base
Affect Performance Team
|News|Nov 20, 2026

Sellforte October 2026: How Test Prioritizer Uses ICE Scores and MMM Data to Rank Incrementality Tests

Sellforte Introduces Test Prioritizer to Identify the Most Important Ad Incrementality Experiments

On October 28, 2026, Sellforte introduced Test Prioritizer, a tool for planning marketing experiments within its Incrementality Testing platform. It analyzes ad spend, past tests, and available measurement methods to determine which campaigns and channels to test first. Test Prioritizer uses the Impact, Confidence, and Ease (ICE) scoring framework to rank potential experiments on a scale from 0 to 10. For companies using Sellforte Marketing Mix Modeling, the recommendations also draw on model results, including estimates of incremental ROAS.

Why Marketers Need to Prioritize Experiments

Incrementality Testing measures the additional results of advertising by estimating how many sales, orders, or other conversions occurred because of ad exposure rather than simply following it. It uses control groups, geographic experiments, and other causal measurement methods.

A full experiment, however, requires time, ad budget, and suitable testing conditions. Sellforte estimates that a typical incrementality test takes four to eight weeks. A large e-commerce company may be running Google Ads, Meta, TikTok, TV, and other channels in several countries at the same time. Testing every combination of campaign, platform, and geographic market is practically impossible.

Teams usually choose tests based on spend, ad platform recommendations, or their own assumptions. The channel with the highest spend may not need another experiment. If recent studies have already established its effectiveness, another test may yield less new information than testing a channel for which reliable data is still unavailable. Test Prioritizer helps teams plan experiments around where new measurements could inform budget allocation decisions.

ICE Score: Three Criteria for Prioritizing Tests

Each potential test hypothesis receives a score from 0 to 10 based on three criteria. Impact measures the business significance of the advertising activity. The amount of ad spend is one of the main factors in assessing the importance of the activity being tested. If a company allocates a substantial budget to a channel, misjudging its effectiveness can lead to significant financial losses. For Sellforte MMM customers, the system can also use incremental ROAS estimates to identify unusually high- or low-performing activities.

Confidence indicates how complete the existing evidence is. This indicates how well the incrementality of a particular channel or campaign is understood. The system considers how many experiments have already been run and how long it has been since the last measurements. For example, a channel that has never been tested may rank high because there is no validated data.

Here, Confidence should not be confused with the statistical reliability of a completed experiment. In Test Prioritizer, the criterion indicates the need for more evidence rather than the likelihood of a positive result.

Ease describes how feasible the experiment is in practice. This assesses how easy it is to set up an experiment using the available methods. For example, the ability to run Conversion Lift on a given ad platform may result in a higher Ease score than a scenario that requires a more complex geographic experiment. Each criterion comes with an explanation, so marketers can see both the overall score and how it was determined.

Ranking Experiments by Channel and Geographic Market

Test Prioritizer displays potential experiments in a table sorted by overall ICE Score. Each row shows the ad channel, market, activity to be tested, and overall priority. For example, Sellforte's demo shows the following results:

  • Meta Sales, Germany: 9.6 out of 10.
  • Meta Sales, United States: 9.2.
  • Meta Prospecting, Germany: 9.2.
  • Meta Advantage+, United States: 9.1.

Three factors contribute to the 9.6 score for the Meta Sales experiment in Germany: substantial ad spend, no prior testing, and the option to run Conversion Lift. These figures illustrate the interface and do not reflect the actual performance of the campaigns shown.

A high ICE Score does not imply high incremental ROAS. It indicates that testing the activity could be especially useful in making future decisions. For example, a high-priority experiment could confirm that an ad channel delivers little incremental lift. That result could help the company cut ineffective ad spend.

Tracking How Much Ad Spend Experiments Cover

Another Test Prioritizer feature assesses how much ad spend is already backed by findings from recent experiments. The interface shows two metrics.

  • The share of ad spend covered by tests in the past 12 months.
  • The number of tested advertising activities relative to the total number of activities in the media mix.

For example, Sellforte's demo account shows that experiments covered 29% of ad spend and 8 of 65 advertising activities over the previous year. These metrics help teams identify gaps in their measurement program.

Ad Spend Covered by Tests
29%

Of ad spend in the past 12 months

Advertising Activities Tested
8 of 65

Across the demo account’s media mix

A company may find that it regularly tests Meta Ads but rarely tests other channels despite spending heavily on them. It may also find that a large share of its budget goes to campaigns whose effectiveness was last measured several years ago.

Test coverage and the accuracy of incrementality measurement are separate matters. A single poorly designed experiment cannot provide a reliable estimate just because it covered a large budget.

Filters for Markets, Ad Platforms, and Testing Methods

Filters can narrow the rankings. Test Prioritizer supports the following filters.

  • A geographic market.
  • An ad platform.
  • A proposed experiment method.

For example, a regional team can select Germany, Meta Ads, and Conversion Lift to see only the test opportunities it is responsible for. Another use is to plan next quarter's experiments for a specific ad channel. After filtering, the team can compare the recommendations with its campaign calendar, seasonal promotions, expected budget changes, and available resources.

Test Prioritizer does not yet run experiments automatically. The tool helps set priorities. Specialists then need to agree on the hypothesis, primary KPI, measurement method, treatment, control group, and test duration.

Using Past Experiments and MMM Results

Test Prioritizer works with Sellforte's experiment library. The library can store results from GeoLift, Meta Conversion Lift Studies, and A/B tests, including the test period, spend, incremental sales, and uncertainty around the estimates. Information from past studies helps identify measurement gaps.

For example, a company may have run several Meta Prospecting experiments in the United States and none in Germany. Even on the same ad platform, results can differ by market because of differences in consumer demand, competition, and campaign structure. Test Prioritizer can account for those differences in its next recommendations.

If the company also uses Sellforte Marketing Mix Modeling, the model provides additional estimates of each channel's incremental effectiveness. This helps identify where an experiment could be useful for checking MMM results. For example, if the model shows unusually high incremental ROAS for a channel, a test can help confirm or revise the estimate.

The feedback also works in the other direction. Results from well-run experiments can be used to calibrate MMM. Sellforte thus connects research planning, experimental data collection, and refinement of marketing effectiveness models.

Test Designer: The Next Step in Automating Experiments

In its October 28 announcement, Sellforte outlined what comes next for the product. Test Designer is scheduled to launch in early 2027. The tool is expected to help turn selected hypotheses into specific experiment plans. Users are expected to receive recommendations on methodology, settings, and testing conditions.

Test Designer was not yet available when Test Prioritizer launched. The Design Test button shown in the interface refers to the upcoming product. These are two distinct stages. Test Prioritizer already helps teams identify which studies to run, while teams continue to design and run tests through existing Incrementality Testing workflows. Sellforte announced a phased rollout of Test Prioritizer to existing customers beginning October 28, 2026.

How to Use Test Prioritizer in E-commerce, Consumer Services, and B2B

For e-commerce companies, the clearest use case is prioritizing which ad channels to test when a company spends heavily on Google Shopping, Meta Ads, TikTok, and other platforms. For example, an online retailer may find that experiments cover its retargeting spend well, while it has never tested prospecting campaigns in some countries. Test Prioritizer helps rank that experiment ahead of retesting campaigns the retailer has already studied.

In consumer services, companies that acquire customers through paid search and paid social campaigns can use the tool to choose where to test the incremental effect of advertising on inquiries or sales. Teams need to define the right outcome for the experiment. More inquiries do not always lead to more qualified leads or paid services.

In B2B, companies with long sales cycles may benefit from prioritizing experiments when they have enough data and a suitable methodology. Low deal volumes, long delays between ad exposure and a sale, and difficulty setting up control groups limit where individual tests can be used. For that reason, ICE Score alone cannot determine whether to run a B2B experiment.

Sellforte Uses Causal Measurement to Inform Ad Budget Decisions

Sellforte is an analytics platform for companies that need to understand how marketing investments affect sales and which budget changes could improve business performance. The product focuses on measuring the actual contribution of ad channels to business results beyond standalone click and conversion metrics. Sellforte uses Marketing Mix Modeling, experiments, and media budget planning tools for this purpose.

Marketing Mix Modeling estimates the contribution of different channels while accounting for seasonality, prices, promotions, and other factors that influence demand. Incrementality Testing adds experimental evidence by measuring how sales change when ad exposure increases, decreases, or stops. Teams can use these measurements to forecast the effects of reallocating budgets and assess the effectiveness of marketing investments.

The platform is designed primarily for e-commerce companies, DTC brands, and retailers that use multiple ad channels and have enough commercial data. For these companies, the challenge is that reporting differs across ad accounts, leaving them unsure how much advertising actually affects sales even when they have plenty of ad data. Sellforte lets teams assess marketing investments in a single model of business performance and test its conclusions through independent experiments.

Test Prioritizer adds a way to manage research priorities. Companies can assess the results of past experiments and plan where new measurements would provide the most useful information for future budget decisions. Incrementality Testing thus becomes a regular way to evaluate the media mix, with individual studies conducted as part of a shared program.