Sellforte AI: Your new colleague for media spend optimization
Ask questions about your marketing performance. Get clear recommendations for where to invest next.
Review marketing performance with Sellforte AI
Understand what drove the result
Ask what changed last week, last month, or year on year. Sellforte AI reviews historical Marketing Mix Modeling results, including spend, incremental sales, ROI and ROAS, and the contribution of baseline demand and promotions.
Break the answer down by channel, platform, region, or campaign to see which activities drove the change and which delivered the strongest return.
Turn the review into a clearer next question
Follow up on a channel, compare periods, or ask for a chart or table. Explore promotional performance by promotion type, discount level, product category, country, customer segment, or sales channel.
Use the model results behind the answer to prepare your performance review and decide where to investigate next.
Media spend of $4.08M drove $12.17M in modeled incremental sales, with an overall iROAS of 2.99.
Amazon Sponsored Products contributed the most at $1.95M, followed by Google Search Generic ($1.47M) and Meta Prospecting ($1.31M). Email had the highest average iROAS at 8.19, on a much smaller budget.
DOOH returned 0.87. I would investigate its campaign and geographic breakdown, then check the stronger channels for room to scale. Use marginal returns and an optimizer scenario before increasing spend; average iROAS alone does not establish the best next investment.
Plan budgets and optimize spend with Sellforte AI
Compare your current plan with an optimized allocation
Ask Sellforte AI to create a budget scenario for the period you are planning. Compare your current allocation with an optimized or forecast plan, then review spend, incremental outcomes, and ROI or CPA by channel.
Keep the total budget fixed to explore reallocation, or compare scenarios with different investment levels.
Understand the trade-offs before committing spend
Compare scenarios side by side to see how a different channel mix changes the modeled outcome. Use marginal incremental ROAS (miROAS) to understand the return on additional spend as channels approach saturation.
Use forecasts to inform your decision, with the planning period, budget, and model assumptions visible in the comparison.
At approximately $4.68M of media spend, the optimized plan projects $17.89M in incremental sales versus $17.22M for the current allocation: about $674,000 more (+3.9%). iROAS rises from 3.68 to 3.83.
The largest increases go to Google Search Generic (+$141,928), Amazon Sponsored Products (+$141,040) and Google Shopping (+$72,030). The main reductions are Brand Search, TV and DOOH.
Analyze and plan incrementality tests with Sellforte AI
Analyze what your experiments measured
Find, summarize, rank, and chart geo tests, Conversion Lift Studies, and randomized A/B tests. Ask what a test measured, which channels or markets it covered, and what the results mean for your next decision.
Review incremental sales and test iROAS alongside uncertainty, the treatment period, and study scope. Compare results with your MMM for the same platform and period to understand where the evidence agrees and where questions remain.
Plan and prioritize your next tests
Ask where your measurement has gaps and which experiment would help resolve an upcoming budget decision. Use past studies to identify channels, markets, or activities that need stronger evidence.
Compare testing opportunities by potential business impact, confidence in the existing evidence, and practical feasibility. Turn the priority into a clear test question and brief for your team. Explore Sellforte Incrementality Testing.
The latest Meta Retargeting study estimates an iROAS of 7.11 (90% interval: 5.83–8.39). Earlier Prospecting and full-funnel studies returned 3.08 and 3.14. Each result applies to its own tested audience, spend and period.
Video Prospecting remains inconclusive at 63% reported confidence. The email winback test showed strong lift, but its small budget makes it a different comparison from paid media.
I would prioritize a Google-only geo holdout: the existing geo studies combine Meta and Google, so they do not isolate Google’s effect. A higher-spend retargeting test could also show whether its strong return holds as investment grows.
Before designing either test, review the detailed studies, control regions, detectable effect and business decision. Keep the planned TV test in the testing roadmap.
Optimize campaigns and ad sets with Sellforte AI
Review performance at the level you buy media
Look back at campaign and ad set performance by platform, channel, account, or campaign. Review period totals or daily trends to understand where spend and outcomes changed.
Move from a high-level channel question into the campaigns behind it, then compare their incremental contribution.
Turn the analysis into bid and target recommendations
Review recommendations for bid parameters within an individual channel or platform. Sellforte AI can recommend bid and target ROAS or CPA changes and explain which campaigns to raise, cut, pause, or adjust.
See the recommendation in the context of the campaign’s performance and modeled opportunity before making a buying decision.
Review US_Shopping_All first: the recommendation lowers target ROAS from 4.85 to 4.03, with projected daily spend rising from $8,285 to $10,828. US_YouTube_Awareness also has room to scale, with miROAS of 3.67 and a suggested target CPA of $20 → $26.33.
For Meta Prospecting, the Broad_Purchase ad set has a smaller recommended increase: daily budget of $6,200 → $6,621. The full recommendation set reallocates approximately $159,873 of daily spend, with major reductions in Brand Search.
The higher target CPA is the recommended bidding setting for the larger allocation. YouTube Awareness spend is projected to rise from about $961 to $1,423 per day, adding approximately $1,837 in daily incremental sales.
Let agents run your scheduled marketing tasks
Schedule the analysis your team repeats
Turn recurring requests into scheduled tasks. Set up a weekly channel review, a monthly performance summary, or a regular review of campaign recommendations, with the request and schedule defined in Sellforte.
Give the agent a specific brief: the period to analyze, the metrics to compare, and the questions the output should answer.
Review the output and keep the task useful
Open completed runs to review the findings and supporting results. Update the request or schedule as your team’s priorities change, or pause a task when it is no longer needed.
Build a repeatable review routine around the analysis, while your team decides which recommendations to act on.
Your recurring marketing updates, ready when you need them.
- Campaign recommendationsWeekly · Mon 08:00Active
- Monthly reviewMonthly · 1st 08:00Active
- Weekly channel analysisWeekly · Mon 08:00Active
Bring Sellforte into your own AI workflows with MCP
Which channels drove extra sales?
Amazon drove the most incremental sales in the reporting period.
What should we test next?
A Google-only geo test would fill a gap in the experiment portfolio.
What if we add 10% to the budget?
I’ll compare incremental sales and iROAS with the original budget.
Where should the next euro go?
Let’s compare a higher-budget scenario with the current plan.
What sales could this plan deliver?
I’ll separate media-driven sales from base demand and promotions.
Turn these results into a presentation.
Here’s a slide outline with findings, assumptions and next steps.
How should we split spend by market?
Let’s compare regional allocations at the same total budget.
Use Sellforte insights in Claude or ChatGPT
Connect the AI tools your team already uses to Sellforte through the Model Context Protocol (MCP). Ask questions from Claude or ChatGPT with access to Sellforte’s measurement and optimization capabilities.
Bring model-backed marketing context into the workspace where you prepare plans, investigate performance, or draft a readout.
Connect analysis to the work around it
Use Sellforte context in a broader AI workflow: prepare a performance report, compare budget options, or investigate experiment evidence alongside your team’s other material.
Explore the connection and available capabilities on the Sellforte MCP page.
Meet the agents powering Sellforte AI
Incrementality
Testing Agent
Understand what worked and decide what to test next.
Media Planner
Agent
Find the media mix that makes your budget work harder.
Media Buyer
Agent
Put your budget and bidding recommendations into action.
Frequently asked questions about Sellforte AI
What is Sellforte AI?
Sellforte AI lets you work with Sellforte through natural-language questions. Use it to review marketing performance, compare budget scenarios, investigate campaign recommendations, and analyze incrementality tests.
What are Sellforte AI’s answers based on?
Sellforte AI uses the data and measurement capabilities available in your Sellforte workspace, including Marketing Mix Modeling, incrementality experiments, and optimization results. The relevant period, scope, and model assumptions matter when interpreting an answer.
Can I use Sellforte AI for performance reviews?
Yes. Ask what changed between periods, which channels drove incremental sales, or how spend and return compare. Follow up by platform, region, campaign, or ad set, and ask for charts or tables to support your review.
Can Sellforte AI help plan and optimize my media budget?
Yes. Compare your current plan with an optimized allocation, explore different total budgets, and review projected incremental sales and returns by channel. Scenario outputs are model-based forecasts; review the assumptions and business constraints before deciding how to invest.
Can Sellforte AI recommend campaign and ad-set changes?
Yes. Review campaign and ad-set performance alongside recommendations for budgets, bids, and target ROAS or CPA. Ask why a change is recommended and examine the modeled opportunity before deciding what to apply.
How does Sellforte AI help with incrementality testing?
Find and summarize geo tests, Conversion Lift Studies, and randomized A/B tests. Review sales impact, test iROAS, and uncertainty, compare experiment evidence with a matching MMM estimate, and identify evidence gaps to help prioritize your next tests.
Can I schedule recurring analysis with Sellforte AI?
Yes. Agents can run scheduled tasks such as weekly channel reviews, monthly performance summaries, and recurring campaign-recommendation analysis. Define the request and schedule, review completed runs, and update or pause the task as your priorities change.
What is the difference between Sellforte AI and Sellforte MCP?
Sellforte AI is the conversational experience inside Sellforte. Sellforte MCP connects Sellforte’s measurement and optimization capabilities to external AI tools such as Claude or ChatGPT, so you can use Sellforte context in your own AI workflows. Explore Sellforte MCP.
