Loading tutorials…
Loading tutorials…
Microsoft offers six bid strategies. Picking the wrong one is the most common reason new accounts plateau at 50% of their potential ROAS. Here's the right one for each account stage.
Who this is forOwners running Microsoft Ads who suspect the bid strategy is leaving performance on the table. Or operators imported from Google who inherited bid strategies that don't fit Microsoft's smaller auction. Don't change bid strategies casually — each switch triggers a 7-14 day re-learning period.
What you'll need
Step 1
Manual CPC, Enhanced CPC, Maximize Clicks, Maximize Conversions, Target CPA, Target Impression Share. Each has data requirements and use cases.
Manual CPC — you set bids per keyword. Most control, most work. Use when conversion data is too thin for any auto-bidding.
Enhanced CPC (eCPC) — you set base bids, Microsoft adjusts ±30% in real-time based on conversion likelihood. Light touch automation, low data requirement.
Maximize Clicks — Microsoft spends your budget to maximize click volume. No conversion-optimization signal. Use for awareness phases, not for ROAS work.
Maximize Conversions — Microsoft spends the full budget targeting the most conversions. Needs 15+ conv/mo to perform.
Target CPA — Microsoft auto-bids to hit a target CPA. Needs 30+ conv/mo for stable performance.
Target Impression Share — Microsoft bids to reach a target % of available impressions (e.g., 80% in top-of-page position). Best for brand defense, not ROAS optimization.
Step 2
0-10 conv/mo → Manual or Enhanced CPC. 10-30 conv/mo → Maximize Conversions. 30+ conv/mo → Target CPA. Brand campaigns → Target Impression Share.
Brand-new campaign (zero history): Manual CPC for 14 days while you build a baseline. Then evaluate.
0-10 conversions/month per campaign: Enhanced CPC. Low data requirement, modest automation. Good fallback.
10-30 conversions/month: Maximize Conversions. Microsoft has enough data to find conversion patterns; no need for a manual CPA target yet.
30+ conversions/month: Target CPA. You can now set a specific CPA goal and trust Microsoft to hit it. Start the target at 20% ABOVE your current actual CPA.
Brand campaigns (your company name): Target Impression Share at 80-90%. Brand bids defend territory; CPA isn't the goal.
Step 3
Campaigns → click the campaign → Settings → Bid strategy → choose strategy → set parameters → save.
Open Microsoft Ads → Campaigns → click the target campaign → Settings.
Scroll to "Bid strategy." Click the current strategy.
Choose the new strategy from the dropdown. For Target CPA: enter the target CPA. For Maximize Conversions: optional bid cap (usually leave blank).
Save. A "Learning" status badge appears next to the campaign for 7-14 days.
During learning: do NOT change targets, do NOT adjust budgets significantly, do NOT pause and reactivate. Let the algorithm find its baseline.
Step 4
Target CPA: start 20% above current actual CPA. Target Impression Share: start at 70% for non-brand, 90% for brand.
For Target CPA: pull your 30-day actual CPA. Set the target 20% HIGHER. Why? The algorithm needs room to explore — too aggressive a target and it gives up on delivery.
After 14 days of stable performance at the higher target, tighten by 10%. Wait another 14 days. Tighten again.
For Target Impression Share: start at 70-80% for non-brand campaigns. Move to 90% only for brand-defense campaigns where presence matters more than CPA.
For Target Impression Share, also set a max CPC ceiling — without it, the algorithm can bid into untenable territory chasing impression share at any cost.
Step 5
First 7 days: expect 30-50% lower volume than steady state. Days 7-14: volume recovers. Day 14+: assess against the target.
Days 1-7 of learning: spend often drops 30-50% as the algorithm explores. This is expected. Do not raise budget or change strategy yet.
Days 7-14: volume should recover to 70-90% of prior baseline. CPA stabilizes near (but usually slightly above) the target.
Day 14: evaluate. Is CPA within 15% of target? If yes, hold and tighten slowly. If no, the target may be unrealistic.
If by day 21 CPA is still 30%+ above target, the target is wrong. Raise it to a sustainable level. Aggressive targets that the algorithm can't hit cause perma-throttling.
Step 6
Once Target CPA is stable, tighten 10% every 14 days. Never tighten more than 20% at once or you'll trigger another learning phase.
Stable performance (CPA within 10% of target for 14 days) → tighten target by 10%.
Wait 14 days. Confirm new target is stable.
Tighten another 10%. Repeat until volume starts to drop meaningfully — that's your account-level CPA floor.
Going below the floor means trading volume for efficiency. Sometimes that's right; usually it's not. Track conversion volume alongside CPA.
Step 7
If a new bid strategy produces 25%+ worse CPA or 40%+ worse volume after 21 days, revert to the previous strategy.
Reverting is fine — bid strategies are a tool, not a commitment.
Common cause of failed Target CPA: insufficient conversion volume. Drop back to Maximize Conversions for 30 days to build volume.
Common cause of failed Maximize Conversions: very narrow targeting. Open targeting before re-trying.
Document each change: date, strategy, target, observed CPA at 14 and 21 days. Builds account knowledge that compounds.
Common mistakes
Setting Target CPA equal to current CPA
What goes wrong: No headroom for exploration. Algorithm under-delivers, CPA actually rises, you blame the strategy. Lost: 30-50% of expected volume during the learning period.
How to avoid: Always start Target CPA at 20% ABOVE current actual CPA. Tighten by 10% every 14 days only after stable performance at the looser target.
Switching bid strategies more than once per month
What goes wrong: Each switch triggers 7-14 day learning. Switch twice and you spend 30 days in learning. CPA, volume, and ROAS all suffer. 60-90 days of suboptimal performance from compounding restarts.
How to avoid: Once you pick a strategy, commit for 30 days minimum. Adjust the *target* within the strategy rather than switching strategies entirely.
Using Target CPA with under 30 conversions/month
What goes wrong: Algorithm has insufficient data to converge on a stable bid. CPA varies wildly week-to-week. You'd be better off with Maximize Conversions or Enhanced CPC.
How to avoid: Hierarchy: 0-10 conv/mo → Enhanced CPC. 10-30 → Maximize Conversions. 30+ → Target CPA. Don't skip stages.
Combining Smart Bidding with low Max CPC caps
What goes wrong: Manual Max CPC cap of $3 prevents the algorithm from bidding higher in moments when conversion likelihood is high. Smart Bidding's advantage is canceled. You pay for the strategy but lose the benefit.
How to avoid: When using Smart Bidding (Target CPA, Maximize Conversions), remove Max CPC ceilings unless you have a hard business reason. Let the algorithm bid freely within budget.
Target Impression Share without a Max CPC cap
What goes wrong: Algorithm chases impression share at any cost. If a competitor enters the auction at $20 CPC, the algorithm bids $20+ to defend impression share. CPC spike of 300% in days.
How to avoid: Target Impression Share requires a Max CPC ceiling. Set it 50-100% above your historical average CPC. The cap protects you from runaway auctions.
Maximize Clicks for accounts with conversion tracking
What goes wrong: Maximize Clicks ignores conversion data. You spend the full budget on clicks regardless of conversion likelihood. CPA balloons because the algorithm isn't trying to optimize toward conversions at all.
How to avoid: Use Maximize Clicks only for awareness/traffic campaigns or for accounts without conversion tracking. The moment you have 10+ conversions/mo, switch to Enhanced CPC or Maximize Conversions.
Recap
Done — what's next
How to set up Microsoft Ads conversion tracking (UET + offline)
Read the next tutorial
Hand it off
Bid strategy choice is where Microsoft Ads accounts either compound or stall. Specialists keep accounts in the right strategy for their conversion-volume tier, tighten targets gradually, and run weekly diagnostics on the learning status. Most accounts on EverestX run $400-1,200/mo for ongoing bid-strategy management at $14-16/hr.
See ongoing management rates
Rarely. Each change costs 7-14 days of learning. Plan for at most one strategy change per quarter, and only when account-stage conditions clearly justify it (crossing a conversion-volume threshold, for example).
Maximize Conversions spends the full budget to get as many conversions as possible — no CPA constraint. Target CPA spends to hit a specific CPA, possibly leaving budget unspent if it can't find conversions at the target. Maximize Conversions is more aggressive; Target CPA more controlled.
If you have 30+ conv/mo and conversion tracking is solid, yes — Target CPA almost always outperforms Manual CPC at that volume. Expect 15-25% better CPA at the same volume after the learning period.
Absolutely — and you should. Brand campaigns: Target Impression Share. High-volume non-brand: Target CPA. Low-volume experimental: Manual or Enhanced CPC. Match strategy to campaign role.
Microsoft is collecting baseline data to calibrate the new bid strategy. During learning, spend may drop 30-50%, CPA may be higher than steady state. Don't change anything during learning. After 7-14 days, status moves to 'Optimizing' or 'Limited.'
Not necessarily. Microsoft's auction is smaller — your conversion volume per campaign is usually 30-50% of Google's. If your Google campaign is on Target CPA (30+ conv/mo), the Microsoft equivalent may only have 10-15 conv/mo and need Maximize Conversions instead.
Microsoft Ads
Without working conversion tracking, every other decision in Microsoft Ads is built on guesses. This walks through UET-based goals AND offline conversion uploads — the path most DIY guides skip entirely.
Microsoft Ads
Microsoft Ads already runs 30-50% cheaper CPC than Google in most categories. With the right tactics, you can drop it another 20-35%. Some take an hour. Some take a month. All matter.
Microsoft Ads
Your Microsoft Ads campaign says "Eligible," budget is set, bids look fine — but impressions are flat. Here's the diagnostic sequence specialists run, in priority order.
Google Ads
Your campaign is enabled, budget is set, and Google says "Eligible." But spend is at 5% of daily budget and impressions are flat. Here's the diagnostic sequence specialists run.