Kuwalyst Media Planner
Build a plan from your data or audited benchmarks, monitor it against your goals and work through recommendations as new evidence becomes available. The source behind each number remains visible throughout.
Media planning software should do three things. It should build the plan from evidence, whether that is your data warehouse or an audited benchmark, and tag every figure as measured, projected or benchmark. It should keep the plan current: monitor it against its goals while the campaigns run, and say in plain words how each euro is contributing this week. And it should argue for every change it proposes, with the hypothesis, the expected impact and a confidence rating, so your team can accept, defer or reject it and keep the record.
Media planning tools that stop at the first step, and marketing analytics solutions that stop at reporting, leave the rest to the planner. The Media Planner is a media planning platform that connects the three: Plan, Operate and Tune in one workspace, with cross platform measurement for online and offline channels alike and, with a warehouse and an MMM, spend optimization that follows marginal incremental ROAS rather than an average return. The marketing mix modeling software inside it runs a Bayesian MMM whose model card, priors and response curves your team can review: how it works is in the marketing mix modeling guide and the marginal incremental ROAS method sheet.
It is built for two kinds of clients. Agencies license it for their client portfolio, and plans, dashboards and presentations carry the agency's brand, so the planners can point to the evidence behind each figure. Brands run it in-house or choose a fully managed service: Kuwalyst's marketing scientists build the plan and run the campaigns, and can also buy the media on request. Among marketing analytics companies and marketing analytics consulting firms, that makes Kuwalyst a marketing analytics company with a product and, when you want it, a marketing analytics consulting partner: either way, the plan, the model and the evidence stay in your workspace. If you are shortlisting the best MMM solutions for retail brands, look at the data-backed start: it connects to the warehouse and the e-commerce platform, and the model accounts for price, promotions and seasonality. The planner connects to your ad accounts and applies changes there directly; by default, each change waits for your one-click approval. As marketing effectiveness software, its job is to show the evidence behind the plan, and to say where your data goes, and where it stays.
Follow the planning process from what is working this week and what to do next, through to your commitments, the portfolio they belong to and the data and models that support them.
Review each channel's committed spend against plan, marginal incremental ROAS and current status. Goals sit alongside the channel view, identified as projected or measured and shown with their target and trend. Weekly and monthly reports explain what changed and where attention is needed.
Each recommendation explains its hypothesis, proposed action, expected impact and confidence rating. You can accept, defer or reject it, and the planner records the decision. Where evidence is limited, measurement recommendations propose an experiment that can replace a modelled figure with a measured one.
Compare allocations side by side, with each option showing its projected blended return, uncertainty range, trade-offs and the guardrails it does or does not meet. Once you commit to a scenario, the planner prepares the plan document, rationale and presentation export.
Designed for teams managing more than one brand, the portfolio shows the evidence available for each one: a calibrated MMM, platform synchronisation or an active A/B test. Plans that need attention are easy to identify.
The planner connects to your data, audited benchmarks, documents and models across online and offline channels. Its connectors cover warehouses, ad platforms, analytics, CRM, e-commerce and flat files, and it can also connect to MCP servers and agent skills developed by your team.
A chat sidebar is open on every screen. Anyone on the team can ask about the plan, pacing or an experiment, and the answers rest on the plan’s figures.
Ask about what is on screen: why the plan favours a channel, which channels are behind their planned spend this week, what the last experiment showed. Answers quote the plan’s figures, each tagged projected or measured. When the assistant proposes a change, it drafts a new scenario or a revision of the plan document.
The assistant answers in the language your team works in, and the plans, reports and alerts it writes are in that language too. Teams in different countries can each work in their own language on the same platform. Figures, channel and campaign names stay exactly as they appear in your data.
You decide which actions the planner can take; by default, it waits for your one-click approval. Each level of autonomy operates within the guardrails and limits you set.
The planner can take low-risk actions within limits you set, such as spend caps and reversal windows, while keeping you informed.
ExamplePausing a keyword with zero conversions.
The default setting. Medium-risk actions are presented for review and can be approved in one click, with thresholds you control for each channel.
ExampleA budget reallocation the MMM recommends.
High-risk actions come with a full briefing before you decide. On higher plans, a Kuwalyst marketing scientist can join the review.
ExampleStopping a campaign or changing a bidding strategy.
If first-party data is not yet available, the plan can be based on your brief and the audited benchmark registry. Each figure is identified clearly, along with the most useful tests to run first.
With a data warehouse and an MMM, allocations can follow the marginal incremental ROAS shown by the model's response curves, rather than relying on an average return.
Your AI analyst inside Shopify admin: what changed in sales, products, inventory and customers, where it came from, and what to check next. In early access, free.
The planner runs open-weight models on Scaleway Generative APIs, hosted in the EU (France): the model's publisher never sees your data. Privately hosted models are available on request, including on-premise and VPC deployments. Read where your data goes, and where it stays.
Starter, Growth and Scale each include onboarding, a monthly review with a marketing scientist and automated execution under human oversight. A management fee on managed media spend is charged by bracket, at lower rates on higher brackets. You can start with a free Media Planner preview.
See pricingBoth, on the same platform. License the Media Planner for your own team, or work with Kuwalyst's marketing scientists on optimisation and execution; agencies license one workspace for their client portfolio, with plans, dashboards and presentations carrying their brand. Whichever option you choose, you use the same workspace with the same visibility into evidence and assumptions, and you can move between the options without losing the data and connections already in place.
The planning and reporting work teams usually do in a spreadsheet and a deck: comparing scenarios, monitoring the plan against its goals, and writing the weekly and monthly reports from the computed figures. It does not replace the ad platforms: it connects to your accounts and applies changes there, with one-click approval by default.
Data warehouses (Snowflake, Google BigQuery, Amazon Redshift), paid media (Google Ads, Meta Ads, LinkedIn Ads, TikTok Ads), web and app analytics (Google Analytics, Mixpanel), CRM (Salesforce, HubSpot), e-commerce and payments (Shopify, Stripe), email (Mailchimp), and files and local databases (CSV, Excel, Google Sheets, DuckDB). Alongside the data: the audited benchmark registry, your briefs and strategy documents, and MMM artefacts, including models your team already built. It can also connect to MCP servers and agent skills developed by your team.
Your workspace runs on Scaleway, in France. The standard AI service is Scaleway Generative APIs, hosted in France, and privately hosted models are available on request, including on-premise and VPC deployments. Data-handling terms, including a right to audit and deletion on request, are set out in the contract. The trust page says where your data goes, and where it stays.
A preview built on market benchmarks is free. The paid plans, Starter, Growth and Scale, differ by the number of brands (one, up to three, up to ten) and the hours of data science support; each adds a management fee on managed media spend, calculated progressively by bracket, with a 4% rate in the first bracket and rates starting at 0.5% in the highest. Annual billing saves 15% on the platform fee, and prices exclude VAT. The pricing page has the figures in euros, dollars, pounds and yen.
“TikTok isn't a channel in your warehouse — you don't have TikTok spend or revenue tracked. There's nothing in the warehouse to calculate a ROAS from right now.”
Begin with audited benchmarks when data is not yet available, then move to a data-backed plan as evidence builds.
Each traced to its primary publisher. The sourcing policy is part of the registry.
Figures are identified as measured, projected or benchmarked on every screen and in every export.
Bring a brief, spreadsheet or data warehouse, and we will show you what a first plan could look like within an hour.
Talk to us