What are SMART Strategies?
SMART Strategies are ready-made Segment Prioritisation strategies that decide which products to boost and which to reduce for you, based on product performance. Instead of creating and maintaining your own segments, you apply a SMART Strategy to your ad set, and it automatically decides which products get more visibility and which get less.
There are currently two SMART Strategies:
Boost ROAS Strategy: Built on SMART segments (Top Performers and Poor Performers), based on ROI Hunter's best practices and Meta's recommendations.
Predictive Boost Strategy: A newer, more advanced strategy that predicts how each product is likely to perform next week and sets an individual priority for every product, rather than grouping them into two segments.
Both are delivered through Segment Prioritisation and applied the same way - from Recommended Setup or directly from the Segment Prioritisation page.
Why the strategies cannot be edited: this is intentional, not a limitation. Locking the strategy gives ROI Hunter a stable foundation to keep improving the underlying algorithm.
Boost ROAS Strategy
A self-optimising strategy that continuously identifies your strongest and weakest products, groups them into SMART Segments, and uses Segment Prioritisation to influence their visibility in Meta’s delivery.
It appears automatically at the top of the strategy overview in Segment Prioritisation page or on the Recommended Setup page for every eligible catalogue.
Managed and evidence-based: ROI Hunter maintains and improves Boost ROAS Strategy continuously, so performance improves without any action on the client's side. It's been tested across multiple advertisers, with a 75%+ success rate in A/B tests.
Practical constraints: it occupies 1 of the 5 strategy slots available per catalogue (a Meta limit).
Requirements
(same as SMART segments eligibility — see Chapter 3):
At least 100 unique currently active products
At least €1,000 (or market currency equivalent) in Meta DPA spend in the last 21 days
At least €10,000 (or market currency equivalent) in product revenue from all traffic in the last 21 days
Meta Match rate equal or above 65% in the last 21 days
Analytics match rate (revenue) equal or above 65% in the last 21 days
Analytics match rate (product detail views) equal or above 65% in the last 21 days
Nice-to-have Meta tracking rate (optional) equal or above 65% in the last 21 days
Boost ROAS Strategy is our recommended setup for clients whose primary goal is improving ROAS on Meta.
For full mechanics of SMART segments, see:
For strategy-building, see:
For troubleshooting, see:
Predictive Boost Strategy
While Boost ROAS Strategy ranks products on performance already recorded, Predictive Boost Strategy adds a forward-looking layer: it estimates how each product is likely to perform next week, based on its recent trend, price, discount, category and the season and sets its Segment Prioritisation priority accordingly.
That means it can start backing a product as it picks up, and ease off one that's starting to fade, before that shift is obvious in a standard performance report.
It's applied the same way as any other SP strategy and delivered through Segment Prioritisation - not a separate feature area.
Getting access
Unlike Boost ROAS Strategy, which appears automatically for every eligible catalogue, Predictive Boost Strategy isn't switched on by default. On Recommended Setup, request Predictive Boost Strategy for your market - once requested, it can take up to 24 hours to appear in the platform.
What makes Predictive Boost advanced
A sharper signal for Meta - every product gets its own priority instead of being placed into one of two fixed groups, giving Meta more precise guidance than Boost ROAS Strategy’s Top/Poor Performers split.
Earlier response to changing performance - Predictive Boost looks at where each product is heading, not just where it’s been, picking up whether recent performance is accelerating or slowing down.
Fairer prioritisation across your catalogue - products are compared with others in a similar category, price range and point in the season, rather than against one flat market average.
Built on long-term context, responsive to what’s happening now - it learns from at least a year of your market’s history, while giving recent performance the greatest influence when setting today’s priority.
Easy to validate against your current setup - because priority is set per product rather than by segment, there’s no Segments in Campaigns breakdown. Instead, Predictive Boost is evaluated like any other strategy: through an A/B test comparing campaign-level results.
Like Boost ROAS Strategy, Predictive Boost is fully automated and self-adjusting - there's nothing to set up or maintain. You apply it the same way, from Recommended Setup or the Segment Prioritisation page.
Predictive Boost Strategy vs. Boost ROAS Strategy
| Boost ROAS Strategy (SMART Segments) | Predictive Boost Strategy |
Basis for priority | Recorded performance | Estimated next-week performance, based on recent trend. Each product is measured against others in the same category, at a similar price point, and at the same point in the season - not against one flat market average. |
Granularity | Two groups: Top Performers / Poor Performers | Individual, continuous priority per product |
History used | Last 21 days | Learns from at least 1 year of history but weighs recent performance ()last 21 days) most, so priorities stay responsive to change. |
Maturity | Proven baseline, recommended starting point | Our most advanced strategy - newer, tested as a challenger to Boost ROAS Strategy |
🗒️ Note: Boost ROAS Strategy isn't being retired or replaced. It stays the recommended, proven baseline. Predictive Boost is the next step for clients who want to test the most advanced option available.
Eligibility Requirements
Here are the current eligibility requirements (last 21 days):
More than 100 active products
More than 1,000 EUR (or equivalent) in Meta DPA spend
More than 10,000 EUR (or equivalent) in product revenue from all traffic
Analytics Match rate of at least 65%, for both revenue and product detail views
Meta Match rate of at least 65%
New for this strategy — more than 5,000 EUR (or equivalent) in Meta DPA product revenu..
New for this strategy — at least 1 year of historical data available on the market. If this isn't available yet, the panel shows "Not available" with a Contact support option to request it.
How to Test Predictive Boost
As with any SP strategy, we recommend testing rather than switching over outright:
Set up the test yourself on the Experiments page.
Test Predictive Boost against the Boost ROAS strategy, or against BAU.
Run the test for 3–4 weeks, aiming for at least 50 conversions per cell.
FAQ
Does Predictive Boost replace Boost ROAS Strategy? Not yet. Boost ROAS Strategy stays available as the safe, proven option; Predictive Boost is a newer strategy to test alongside it.
My market meets Boost ROAS Strategy's requirements - why I cannot request Predictive Boost? Predictive Boost has one extra requirement: at least 1 year of historical account data on the market. If you believe your market qualifies and it's still missing, contact you CSM.
Where can I see the impact of Predictive Boost? Through campaign-level results in an A/B test. Because it sets an individual priority per product rather than by segment, there's no segment-level breakdown for it the way there is for Boost ROAS Strategy.
What is Boost ROAS Strategy, and how is it different from building the same segments myself? Boost ROAS Strategy is pre-configured and locked; ROI Hunter actively manages it and rolls in algorithm improvements automatically. A custom strategy using the same SMART Top/Poor Performers segments won't receive those updates on its own.