Scaling and diagnostics
Growing budget safely, and finding the cause systematically when performance drops.
Scaling is not a budget operation, it is a capacity question: can your current creative and bid setup absorb more money and still produce profit? If it cannot, raising budget only lowers efficiency.
Are you ready to scale
Four conditions have to hold at the same time before you raise budget:
| Condition | Threshold |
|---|---|
| Signal | The ad set clears ~50 conversions per week |
| Profitability | MER is above break-even and has held for 2 weeks |
| Creative depth | 3 or more healthy creatives running at once |
| Production line | New creative is ready for the next 2 weeks |
The third and fourth are the most commonly skipped. Scale an account that rests on a single creative and the whole account drops the day that creative fatigues.
How to scale
| Method | Practice | Risk |
|---|---|---|
| Vertical | Raise the existing campaign's budget 20–30%, wait 2–3 days | Low |
| Horizontal (creative) | Add new creatives to the same ad set | Low |
| Horizontal (placement/geo) | Open a new country, language or placement | Medium |
| New channel | Expand to another platform | High |
That order is the recommended order. Opening a new channel is the most expensive move: a new learning phase, a new creative format, and a new measurement setup. Do not go there before you have hit the ceiling on your current channel.
Do not apply more than one scaling method in the same week. When results change, you will not be able to tell which one caused it.
When performance drops: diagnostic order
Reacting to a drop in a panic — pausing everything, cutting budget, rebuilding from scratch — usually makes it worse. Eliminate in order:
1. Did measurement break? Check the pixel and server-side event flow. If order count is flat while reported conversions fell, the problem is tracking, not sales.
2. Did something external change? Season, holidays, a stockout, a price change, a new site bug, a competitor's campaign. Causes outside the ad account are more common than causes inside it.
3. Is it the site? If CTR is flat and CVR dropped, the problem is not the ad. Check landing page speed, checkout flow, and inventory.
4. Is it the creative? Look at hook rate and frequency. See creative fatigue for the full table.
5. Is it delivery? Review learning phase status and any edits made in the last 7 days. Most sudden swings trace back to an edit made a few days earlier.
6. Is it the auction? If CPM rose sharply and everything else is flat, competition increased. This is usually seasonal, and your only real lever is creative quality.
Quick diagnostic table
| Symptom | Most likely cause |
|---|---|
| CPM flat, CTR down | Creative fatigue |
| CPM up, CTR flat | Auction competition or too narrow an audience |
| CTR flat, CVR down | Site, inventory or price |
| Everything flat, reported conversions down | Tracking or attribution |
| Results swinging day to day | Learning phase unfinished, signal too sparse |
| Spend falling short of budget | Bid cap set too low |
Channel mix
When running multiple channels, allocating budget by platform ROAS is a systematic error — bottom-funnel channels always look good and pull budget toward themselves. Allocate on three inputs instead:
- Incremental contribution — the real effect from a blackout or geo test
- Total MER — at the business level, not per channel
- Saturation — can the channel absorb more budget without losing efficiency
When all three agree, the decision is easy. When they do not, the incrementality test — not the platform dashboard — should decide which channel grows.