Recast vs Sellforte (2026): Channel vs Campaign-Level MMM
Recast and Sellforte both build Bayesian marketing mix models (MMM) calibrated with incrementality tests. The main differences: Sellforte measures and recommends at campaign and ad set level, updates results daily and is built for ecommerce and retail. Recast measures at channel and tactic level, refreshes its models weekly and serves a broader set of industries, including CPG, fintech and pharma.
Reviewed November 2026. Sellforte is our product. Every Recast fact comes from Recast's public pages, listed under Sources. Corrective feedback is welcome at research@sellforte.com.
Recast is a Bayesian marketing mix modeling platform. It offers Recast MMM, GeoLift by Recast for geo experiments, forecasting and planning tools and a budget optimizer. It works with Fortune 500, CPG, fintech, pharmaceutical and DTC brands, including Canva, Block, Harry's, PODS and GetYourGuide.
Sellforte is a marketing measurement and optimization platform for retail and ecommerce brands. It combines causal MMM, incrementality testing and incrementality-corrected attribution, and turns the results into campaign and ad set level recommendations. Customers include C&A, Lidl, bonprix, Douglas, Tchibo and INTERSPORT.
How is Sellforte different?
1. We measure and recommend at campaign and ad set level
Sellforte calculates incremental ROAS and marginal incremental ROAS (the return on the next dollar) for every campaign and ad set, and turns them into budget and bid recommendations at the same level. Approve a recommendation and Sellforte pushes it to Meta, Google or TikTok.
Recast measures at channel and tactic level. For major channels such as Meta and Google, it usually splits the channel into tactics, such as prospecting and retargeting. Campaigns and other sub-channels get performance indices that rank them within their parent channel. Recast's documentation says these indices are not exact efficiency estimates, are not time-varying and should be read directionally.
In practice: With Sellforte, a media buyer can see which Meta ad sets to scale and which to cut in incremental terms, without switching to platform-reported numbers or multi-touch attribution for decisions inside a channel. See it with demo data: which campaigns are pulling down a channel's return and which campaigns and ad sets are worth scaling.
2. We update results every day
Sellforte updates incremental ROAS at campaign and ad set level every day. Results update with the daily data pipeline, so a campaign that ended last week already has modeled results. That matches how performance teams work: budgets and bids change during the week, and a weekly read can arrive after the decision has already been made.
Recast refreshes its models weekly, with on-demand refreshes available.
3. We specialize in ecommerce and retail
Sellforte is built for retail and ecommerce brands, such as fashion, grocery, beauty, DIY, furniture and DTC. It models marketing impact on every sales channel a retailer has: stores, app, ecommerce, marketplaces and wholesale. It also has retailer-specific features, such as measuring the incremental sales and margin impact of promotions by discount depth, promotion type and category.
Recast's customer reference list suggests a generalist industry focus that includes retail and DTC, but also CPG/FMCG, fintech and pharma.
See incremental ROAS for every campaign and ad set
We'll walk you through Sellforte with campaigns like yours: daily results, budget and bid recommendations, and the tests that calibrate the model.
Comparing Recast and Sellforte
The tables below compare the two platforms area by area. "Not published" means we could not find the fact on the vendor's public pages.
Core features
Both platforms are built on Bayesian MMM calibrated with incrementality tests, and both offer forecasting, budget optimization and an MCP server. The differences are in depth and speed: Sellforte measures down to campaigns and ad sets, updates daily, adds attribution and pushes approved changes to ad platforms. Recast measures at channel and tactic level and refreshes weekly.
| Feature | Sellforte | Recast |
|---|---|---|
| Marketing mix modeling | Bayesian MMM | Bayesian MMM |
| MMM calibrated with incrementality tests | ✓ YesSee MMM calibration | ✓ Yes |
| Incrementality testing | ✓ YesGeo tests, conversion lift studies and A/B tests | ✓ YesGeo tests (GeoLift by Recast) |
| Multi-touch attribution | ✓ YesMTA based on Incremental Pixel | ✗ No |
| Incrementality-corrected attribution | ✓ YesWith Causal Attribution | ✗ No |
| Lowest level of incremental ROAS | Campaign and ad set | Channel and tacticCampaigns get relative performance indices |
| Results refresh | Daily | Weekly, or on demand |
| Forecasting and budget optimization | ✓ YesWith Optimizer | ✓ Yes |
| Push changes to ad platforms | ✓ YesTo Meta, Google and TikTok after approval | ✗ No |
| Promotion measurement | Incremental sales and margin impact by discount depth, promotion type and category | Included in MMMNo breakdown by discount depth or category published |
| MCP server | ✓ YesSellforte MCP covers every product module | ✓ YesAccess to most reports in Recast's Reporter |
Platform and pricing
Both are software platforms that come with expert support. Sellforte also offers self-serve model configuration and publishes its pricing. Recast's team or a certified partner configures each model. Recast does not publish its MMM prices, and GeoLift by Recast comes with a six-month free trial.
| Feature | Sellforte | Recast |
|---|---|---|
| Delivery model | Software platform with a dedicated Customer Success Manager, training and support in a shared Teams or Slack channel. Self-serve model configuration available for data scientists and marketing analytics teams | Web platform with analyst support, configured by the Recast team or a certified partner |
| Time to first results | 1 to 3 weeks for small and mid-sized ecommerce businesses, 2 to 4 months for larger enterprises | Initial results 2 to 3 weeks after receiving data |
| Pricing | Published: from $1,900 a month for Incrementality Testing, $2,500 for Incremental Attribution, $4,500 for Full-Funnel MMM and $8,500 for Enterprise | MMM: Not publishedGeoLift by Recast: six-month free trial, also included with Recast MMM |
| What sets the price | Package, average monthly media spend and sales channels in scope | Not published |
Channel and tactic level measurement and optimization
These are the traditional MMM use cases, and both platforms cover them: incremental and marginal ROAS, saturation curves, budget allocation and revenue forecasts by channel. If channel-level answers are all your team needs, either platform does the job.
| Feature | Sellforte | Recast |
|---|---|---|
| Incremental ROAS by channel | ✓ Yes | ✓ Yes |
| Marginal incremental ROAS by channel | ✓ Yes | ✓ Yes |
| Saturation curves by channel | ✓ Yes | ✓ Yes |
| Optimal budget allocation by channel | ✓ YesWith Optimizer | ✓ Yes |
| Revenue forecast based on optimized spend | ✓ Yes | ✓ Yes |
Campaign and ad set level spend optimization
This is where the two platforms differ. Sellforte takes the same measurement down to the campaigns and ad sets media buyers manage every day, recommends spend and bids at that level and tracks the impact of each change. Recast estimates incremental ROAS down to channel and tactic level, ranks campaigns with directional indices and can split a channel's budget across them.
| Feature | Sellforte | Recast |
|---|---|---|
| Incremental ROAS by campaign and ad set | ✓ YesWith Sellforte Performance | ✗ NoCampaigns get directional indices that rank them within their parent channel |
| Marginal incremental ROAS by campaign and ad set | ✓ Yes | ✗ No |
| Optimal spend by campaign and ad set | ✓ Yes | PartialSplits the parent channel's budget across sub-channels. Preset blends keep 90 to 98% on historical spend shares |
| Optimal bidding values by campaign and ad set | ✓ Yes | ✗ No |
| Executes bidding changes on ad platforms | ✓ Yes | ✗ No |
| Tracks the impact of bidding changes with campaign and ad set level pre/post analysis | ✓ YesWith Change Intelligence | ✗ No |
Incrementality testing
Both platforms feed experiment results back into the MMM. Recast's own testing tool, GeoLift by Recast, runs geo tests, and its team can add results from in-platform lift studies such as Meta conversion lift. Sellforte Incrementality Testing covers geo tests, conversion lift studies and A/B tests, with every result in one Test Library and a Test Prioritizer that finds the next testing opportunity.
| Feature | Sellforte | Recast |
|---|---|---|
| Geo tests | ✓ YesSellforte Geo Tests | ✓ YesSpend-up, spend-down and go-dark, with synthetic controls |
| Conversion lift studies | ✓ YesConversion Lift Studies in Sellforte | PartialResults of in-platform lift studies, such as Meta conversion lift, are added to the MMM manually by Recast's model building team |
| A/B tests | ✓ YesSellforte A/B Tests | Not published |
| Test design | Test Library, Test Prioritizer and the Incrementality Testing Agent | Power analysis for geo split, budget and detectable lift |
| Results feed the MMM | ✓ YesAs Bayesian priors, see MMM calibration | ✓ Yes |
Model validation and accuracy
Both vendors test their models against data the model has not seen. Forecast accuracy is central to Recast: every customer gets out-of-sample forecast accuracy reports each week, and the platform has built-in backtesting. Recast's model also re-estimates channel ROIs every week, so they can change over time, and models lower-funnel channels such as branded search and affiliates as outcomes of upper-funnel spend. Sellforte tracks out-of-sample accuracy, bias and backtests over time in Model History.
| Feature | Sellforte | Recast |
|---|---|---|
| Out-of-sample forecast accuracy | ✓ YesMAPE and bias tracked over time in Model History | ✓ YesWeekly forecast accuracy reports for every customer |
| Backtests | ✓ YesIn Model History | ✓ YesBuilt-in backtesting |
| Public forecast accuracy reporting | ✗ No | ✓ YesForecasts from every production model are scored weekly, with results on Recast's research page |
Security and compliance
Both have third-party security assurance: Recast reports SOC 2 Type 2 compliance, and Sellforte is regularly audited by Nixu (part of DNV). Sellforte publishes its hosting options and single sign-on support. Recast's trust center lists Google Cloud and AWS among its subprocessors, and Recast supports SAML single sign-on through the customer's identity provider.
| Feature | Sellforte | Recast |
|---|---|---|
| Audit or certification | Regularly audited by Nixu (part of DNV) | SOC 2 Type 2 |
| Data hosting | Option between US and EU | Location not publishedSubprocessors include Google Cloud and AWS |
| Single sign-on | Azure AD and Google SSO | SAML SSO through your identity provider, such as Okta |
Integrations
Both connect to common data warehouses such as BigQuery and AWS. Sellforte adds more than 500 built-in connectors and partnerships with Meta and TikTok. Recast offers public, versioned APIs alongside its MCP server.
| Feature | Sellforte | Recast |
|---|---|---|
| Getting data in | More than 500 built-in connectors, direct data warehouse connections (BigQuery, AWS, Azure and Snowflake), plus data upload for offline media | Programmatic access to daily marketing data, through the client's BigQuery, Google Cloud Storage, AWS S3 or Google Sheets, or Recast's S3 bucket or SFTP server |
| Ad platform partnerships | Meta Business Partner, TikTok Marketing Partner | Meta Measurement Partner |
AI agents and MCP
Both vendors let AI assistants work with the model through the Model Context Protocol (MCP).
- Recast offers an MCP server alongside its APIs, with access to most of the reports in its Reporter (20 at launch in July 2026). It connects to Claude through a one-click plugin, and to ChatGPT, Codex and other MCP clients. Recast says you can ask questions about your model in plain English, blend Recast results with your own data and push results to Slack.
- Sellforte makes every product module available through Sellforte MCP, which works with Claude, ChatGPT, Gemini and other AI assistants. It also includes three AI agents: the Media Planner Agent, Media Buyer Agent and Incrementality Testing Agent. Recommendations can be applied to ad platforms after you approve them.
Industry focus and customer references
Most of Sellforte's public references are retailers and ecommerce brands, alongside restaurant chains, consumer brands, gaming, payments and telecom companies. Recast's references lean toward fintech, DTC and consumer brands, plus software, media, travel and pharma. The table lists the customers each vendor names on its public pages, grouped by segment.
| Segment | Sellforte | Recast |
|---|---|---|
| Fashion, apparel and accessories | C&A, bonprix, Peek & Cloppenburg, KiK, Azzas 2154, Represent, Odlo, Brahmin, ALOHAS, Cris Barros | Negative Underwear |
| Beauty and personal care | Douglas | Harry's, Lume |
| Grocery and food | Lidl, Stockmann Delicatessen, Matsmart | Daily Harvest |
| Consumer electronics | Verkkokauppa.com, Caseking | Not published |
| Home and variety retail | Finnish Design Shop, Tchibo, Tokmanni | Not published |
| Sporting goods, pet and specialty retail | INTERSPORT, Musti Group, Fressnapf, smartphoto | Not published |
| Restaurants | Subway Nordics (Nordic Bites Group), Kotipizza | Not published |
| Consumer brands (CPG) | Fazer, PPG | Hasbro |
| Ticketing, lottery and gaming | eBilet, Svenska Spel | Not published |
| Fintech and payments | Paysafe | Block (Cash App, Square), KOHO, Rocket Money |
| Telecom | Telenor, Telia, Moi Mobiili | Not published |
| Travel | Not published | GetYourGuide |
| Pharma | Not published | Currax |
| Software, media and home services | Not published | Canva, Los Angeles Times, PODS |
When to choose Recast vs Sellforte
Choose Sellforte if
- You are a retail or ecommerce brand selling online, in stores or both.
- Your team optimizes campaigns and ad sets and wants incremental results at that level every day.
- You want geo tests, conversion lift studies and A/B tests in one place.
Choose Recast if
- You are a CPG, fintech, pharma or subscription brand.
- Channel-level answers, refreshed weekly, are enough for how you plan.
- You want forecasts tracked against finance targets with weekly accuracy scorecards.
Recommendations by team type
| Team type | Better fit | Why |
|---|---|---|
| Ecommerce and DTC performance marketing | Sellforte | Daily campaign and ad set results, with recommendations you can push to Meta, Google and TikTok |
| Omnichannel retail | Sellforte | Store, web, app and marketplace sales in one model, with promotions measured by discount depth, promotion type and category |
| Marketing analytics teams needing self-serve calibration and configuration | Sellforte | Interfaces to review and edit modeling assumptions and calibration inputs |
| CPG selling through retailers | Recast | CPG is a core segment. Sellforte is not designed for manufacturers that sell mainly through other retailers |
| Fintech, subscription and apps | Recast | Multi-stage funnel modeling from signup to revenue, with fintech customers such as Block and Rocket Money |
| AI workflows on measurement data | Either | Both offer MCP servers. Sellforte exposes every product module and can apply approved changes to ad platforms |
Frequently asked questions
What is the difference between Recast and Sellforte?
Both build Bayesian marketing mix models calibrated with incrementality tests. Sellforte measures and recommends at campaign and ad set level, updates results daily and specializes in ecommerce and retail. Recast measures at channel and tactic level, refreshes weekly and serves a broader set of industries, including CPG, fintech and pharma.
Is Sellforte a good Recast alternative?
For retail and ecommerce brands that optimize campaigns and ad sets, yes. Sellforte gives daily incremental ROAS at that level and publishes its pricing. For CPG, fintech or subscription businesses, Recast may fit better.
Does Recast measure campaigns and ad sets?
Recast measures channels and, for major channels, tactics such as prospecting and retargeting. Campaigns within a channel are shown as performance indices ranked against their parent channel. Recast's documentation says these indices should be read directionally and are not time-varying. Its optimizer can split a channel's budget across those campaigns.
How often do Recast and Sellforte update results?
Recast refreshes its models weekly, with on-demand refreshes available. Sellforte updates incremental ROAS at campaign and ad set level daily.
How much do Recast and Sellforte cost?
Recast does not publish its MMM pricing, and GeoLift by Recast comes with a six-month free trial. Sellforte's published plans start at $1,900 a month for incrementality testing and $4,500 a month for full-funnel MMM. Prices depend on the package, average monthly media spend and sales channels in scope.
Do Recast and Sellforte offer incrementality testing?
Yes. Recast offers geo tests through GeoLift by Recast. Sellforte offers geo tests, conversion lift studies and A/B tests. Both feed test results back into the MMM.
Do Recast and Sellforte have MCP servers?
Yes. Recast offers an MCP server with access to most of its reports, alongside its public APIs, and it works with Claude, ChatGPT and Codex. Sellforte MCP makes every Sellforte product module available to Claude, ChatGPT, Gemini and other AI assistants.
Considering Sellforte as your Recast alternative?
See campaign and ad set level results on data like yours, and get a price for your media spend and sales channels.
Sources
Every vendor fact in this article comes from these public pages, read in November 2026. Corrective feedback is welcome at research@sellforte.com.
- Recast website: getrecast.com (homepage, including its customer logos and testimonials), getrecast.com/mmm (Recast MMM), the GeoLift by Recast page, getrecast.com/recast-llm-information (company overview), getrecast.com/mcp-is-here (Recast MCP launch post), getrecast.com/product-demos (backtesting and accuracy scorecards), getrecast.com/research (published forecast accuracy), the blog post on combining MTA and MMM, and getrecast.com/recast-case-studies with the case studies for Block, KOHO, LA Times, GetYourGuide, Lume, PODS, Daily Harvest and Negative Underwear
- Recast documentation: docs.getrecast.com (How to use Sub-channels in Recast, When should I split a channel into more granular breakouts, Integrate Recast into your weekly marketing workflow, Data Guide, Data Sources, Multi-stage modeling, Incrementality Experiments, Lift Tests and Incrementality Experiments, Reporter, Onboarding FAQ, The Recast MCP, MCP Setup Guide) and geolift-docs.getrecast.com (Subscription Management, Frequently Asked Questions)
- Recast trust center: trust.getrecast.com (compliance, subprocessors)
- Sellforte: homepage, Pricing, Data Security, Connecting Data, Promotions, Sellforte MCP, Marketing reporting with Sellforte MCP, Incrementality Testing, Model History, Customer cases, Customer News and the industry solution pages
You May Also Like
These Related Stories
Measured vs Sellforte (2026) for MMM, Incrementality testing and attribution

