Announcements ·

Introducing Horizons 2026: Six updates built around one idea

Forecaster 2026 brings six updates to your dashboard: guided onboarding, AI Action Center, Media Forecaster, Attribution, Model Center, and Data Center.

Introducing Horizons 2026: Six updates built around one idea

Marketers have gotten used to a version of measurement that asks for a lot of faith. A dashboard tells you a channel is underperforming, a model recommends a budget shift, and you're left to either trust it or spend your own time trying to verify it. That trade-off has never made much sense, especially when the underlying modeling has been solid the whole time.

Knowing you can trust a number also matters less if you're still spending half your week hunting for it across five different tools to act on it. That's the second gap this release closes. Six updates, one shared idea: more direct control over your model, budget, and data, without the busywork of stitching it all together yourself.

Key takeaways

  • Onboarding is now guided, showing you only the connectors you actually need and flagging issues before they become blockers.
  • The AI Action Center is an in-product AI assistant that answers questions about your performance directly from your own model.
  • Media Forecaster brings planning and forecasting together into a single tool, so you get one plan instead of two conflicting ones.
  • Attribution breaks down how much of your results come from a channel's direct impact versus the halo effects it creates elsewhere.
  • Model Center gives you a clear, direct view into how healthy and accurate each of your models is, and why.
  • Data Center lets you manage your own integrations and connectors without waiting on a ticket.
  • All six updates are live in your dashboard now, at no additional cost.

Know exactly where you stand with guided onboarding

Getting started used to mean staring at a full catalog of connectors, guessing which ones you actually needed, and handing the rest off to a developer with no clear view of what was done or what was still blocking you. It's a rough way to start a relationship with a new tool.

Onboarding is guided now. It shows only the connectors your account actually needs, tracks your progress toward a clear finish line, and flags issues early, so a missing permission or a misconfigured source surfaces before it can hold up your first model. Instead of wondering whether something fell through the cracks, you always know what's done and what's left, and so does anyone else on your team helping get set up.

Ask your model directly with the AI Action Center

"How much did retail really drive last month?" used to mean digging through a dashboard until you found the answer, or waiting on someone else to pull it for you. Neither option was especially fast, and both meant your question sat unanswered longer than it needed to.

The AI Action Center includes an in-product AI assistant that answers straight from your own model, in plain language. Ask for a recap of last month's performance, dig into why your modeled ROAS differs from what a platform is reporting, or compare one channel against another. You get an answer in the time it takes to type the question, instead of the time it takes to build a report. The AI Action Center works alongside your CSM rather than replacing that relationship, and it stays inside your own data the whole time, so the answers you get back are grounded in what your model actually knows.

One plan, not two, with Media Forecaster

Planning and forecasting used to live in separate tools, which sometimes meant two different action plans pointing your budget in two different directions. Media Forecaster brings the two together into one home. Set a single goal, and you get one plan across every model you run, with the reasoning behind each move shown before you commit. Instead of reconciling two tools on your own time, you walk into the meeting room with one number and the ability to defend every line of it.

See where every channel actually earns its budget with Attribution

It's easy to assume your lowest-ROAS channel is your weakest one. Often, it's actually the channel creating demand that shows up somewhere else, like branded search, direct traffic, or your Amazon storefront, and a channel-level number alone won't tell you that.

Attribution makes that visible. The model still breaks down the impact of every channel and campaign so you can see the direct result on its own (Base) alongside the cross-channel lift it's creating elsewhere (Halo). Your modeled total sits right next to what the platform reports, so you can sort it, filter it, and bring it straight into your next planning conversation.

Trust the model because you can see it, with Model Center

A model can be right and still get second-guessed if nobody can explain why it said what it said. That's been the trade-off with MMMs for a long time: powerful modeling, with no easy way for the people using it to see the reasoning behind it, or to explain that reasoning to anyone else.

Model Center changes that. Every model you run lives in one place, along with a Model Accuracy score, a backtest showing how it's performed against real outcomes, and a record of the assumptions and configuration changes behind it. When a recommendation shifts from what it was last month, you can see what actually changed without needing a data science background.

Manage your own data with Data Center

Routine changes to your setup, like mapping a new campaign into your model, used to mean opening a ticket and waiting on engineering to get to it. Nothing about that delay was your fault, but it was still your budget sitting idle in the meantime.

Data Center puts your whole setup on one screen: integrations, connectors, and mapping. Connect a source, validate it, and deploy the change yourself, all without leaving your dashboard. Your model keeps pace with your media plan, and the small changes that used to take days now take minutes.

Wrapping it up...

Six updates, one idea: less time spent piecing together your own answers, more time acting on the ones your model already has. From the moment you get started to the moment you're managing your own data, every part of that experience just got more direct, and none of it requires learning a new system or waiting on someone else to walk you through it.

All six are live in your dashboard now, at no additional cost. Log in and you'll see them waiting for you.

The Halo

Exclusive insights, every week.

Subscribe to The Halo for sharper marketing thinking.

Keep reading