Google's smart bidding algorithm looks at the last 30 days when deciding how and where to spend. That works fine in steady-state — but during a short, high-volume event like Black Friday, it creates a critical lag. By the time the algorithm recognises a spike in conversion rate, you're already 1–2 days into the event and have missed the peak.

The second problem is the post-sale phase: after the sale ends, the algorithm has just "learned" that high spend was efficient, so it continues scaling — burning budget at peak-event CPCs during completely normal traffic. Touchpoint's portfolio manages €80–90M in annual ad spend, and they've found that keeping CPCs stable while growing revenue requires actively working against the algorithm's default behaviour. SBAs are the main tool they use for this.

Seasonality Bid Adjustments (SBAs) are a native Google Ads tool that lets you manually signal a predicted change in conversion rate to the algorithm — independently of its normal 30-day bidding signals. In practice, this lets you:

You find the tool under Tools → Bidding → Adjustments in the Google Ads interface. Always use Conversion Rate adjustment (not budget adjustment) — budgets in Touchpoint's setup are uncapped, so conversion rate is the right lever. Set a timeframe, select your campaigns, and enter your predicted CVR change.

In the context of SBAs, a seasonality is anything that has a positive or negative impact on your shop's conversion rate — and that the algorithm doesn't know about yet, or would only understand with a time lag. This includes:

The key test: does this event have a clear, short-term effect on your CVR that Google wouldn't pick up within the first few hours? If yes, it's worth an SBA.

1. Pre-sale preparation

Look at last year's sale data: what was the actual CVR uplift during the event? How did spend and revenue behave hour by hour? This historical data is what you use to calibrate your initial adjustment. Don't just guess a number — and don't start aggressively. Your first time using SBAs, start conservatively and adjust intraday based on what you see.

2. Setting the bid adjustment

At the moment your sale goes live, set your positive SBA. Always split Brand and Non-brand campaigns — the same scaling logic doesn't apply to both equally. You can get as granular as individual product campaigns if the event is product-specific.

3. Intraday monitoring

This is where the "intraday" part actually happens. On a strong sale day, Touchpoint makes up to 16 bidding changes. They monitor using by-time segments in Google Ads (segment by Hour of Day) and watch conversion value and conversion rate for the current time window — not click-attributed data, but what's actually arriving in the account right now. If CVR is holding strong, they push the adjustment higher. If it's not keeping pace, they pull back.

4. Post-season adjustment (don't skip this)

When your sale ends, immediately set a negative SBA (start at -50% as a benchmark) to prevent the algorithm from continuing to scale at event-level efficiency. Without this, Google will look at its 30-day window, see high performance, and keep spending aggressively — at CPCs that only made sense during the sale. Touchpoint learned this the hard way: they once needed to keep a negative adjustment running for 2.5 weeks before the algorithm normalised. Never overwrite old SBAs — leave them in your account so you can reference them for the next event.

Separate from SBAs, Google's smart bidding algorithm naturally cycles between buffer phases (building efficiency reserves) and scaling phases (spending that reserve down). The problem: in the scaling phase, Google can inflate CPCs significantly — especially on brand campaigns where it knows conversions are easy. Touchpoint has seen brand clicks cost €16 on accounts where the average CPC was set at a much lower target.

The fix is dynamic target setting: small, regular adjustments (±10%) to your target ROAS based on your competitive position:

Applied consistently, this keeps your brand CPC stable even as spend grows. For high-volume brands, the savings potential runs to six figures annually — roughly 10% of brand spend that would otherwise go to inflated auction prices. This works on non-brand campaigns too, but be more cautious there since non-brand conversions are harder for the algorithm to predict accurately.

80–85% of Shopping matching between a search query and a product listing happens in the product title. Most brands use one static title for every search context — the same title whether someone searches a brand term, a solution-aware term, or a problem-aware term.

Touchpoint's approach with one of their clients (AG1 / Athletic Greens): four separate feeds, each with titles optimised for a different search intent cluster:

Result: up to 60% higher CTR on the theme-matched feeds compared to a single generic title. You're effectively using 150 × 4 = 600 characters across four contexts instead of 150 characters for everything.

How to set it up technically: duplicate the item ID (not the product ID — Google treats that as the same product) and assign each duplicate a unique feed label. Google splits data by feed label in the interface, so you get clean per-feed reporting. You do not need new product IDs — using the same item ID across feeds is what allows Google to treat them correctly.

Important constraint: each feed cluster needs at least 30 conversions per 30 days to run its own campaign effectively. Below that threshold, your conversion data is too fragmented to give the algorithm meaningful signal. Start with the clusters that already have enough conversion volume in your main feed.