Enhanced CPC: When and How to Use It

Enhanced CPC is gone. Google fully retired it as a standalone bid strategy through 2025, and by 2026 the option you may still have a muscle-memory click for is not there. If you came here to learn when to turn ECPC on, I have bad news and good news. The bad news: that decision no longer exists. The good news: it was never the decision that mattered.

This is not a "how to configure enhanced CPC" post. That post is obsolete the moment it is published. This is a post about the thing the ECPC deprecation should have made everyone ask and almost nobody did: what is your bidding strategy actually trained on?

Because here is the part the migration guides skip. Whether you run manual CPC, Target CPA, Target ROAS, or whatever Google nudges you toward next quarter, every one of those strategies optimizes against your conversion data. If that data is dirty - stuffed with bot conversions, missing real ones blocked by ad blockers - then the strategy choice is a rounding error. You are tuning the steering wheel while the map is wrong.

DataCops is the architectural answer to the map problem: first-party conversion tracking that filters bots before the signal ever reaches Google. But let me walk the argument first.

Quick stuff people keep asking

What is enhanced CPC in Google Ads? It was a semi-automated bid strategy. You set manual CPC bids and Google adjusted them up or down in real time based on conversion likelihood. A hybrid between manual control and Smart Bidding.

Is enhanced CPC still available in 2026? No. Google deprecated ECPC as a standalone strategy through 2025. In 2026 it is not a selectable bid strategy for search and display in the way it used to be.

What replaced enhanced CPC after deprecation? Google's answer is full Smart Bidding - primarily Target CPA and Target ROAS, with Maximize Conversions and Maximize Conversion Value as the volume-oriented options. For accounts without enough data, manual CPC is still the honest fallback.

When should you use manual CPC bidding? When you do not yet have the conversion volume for Smart Bidding to learn from. New campaigns, low-volume accounts, tight niches. Manual CPC is also the right call when you do not trust your conversion data, because at least manual bidding does not amplify a bad signal automatically.

How many conversions do you need for Smart Bidding to work? The rough industry rule is around 30 conversions in 30 days per campaign as a floor, and more is better. Below that, Smart Bidding is guessing, and it guesses worse the dirtier the data is.

What is the difference between enhanced CPC and Target CPA? ECPC adjusted your manual bids within a limited range around a conversion-likelihood signal. Target CPA hands bidding fully to Google with a cost-per-acquisition goal. Target CPA has more freedom - and more dependence on clean conversion data.

Should small businesses use Smart Bidding or manual CPC? If you are below the conversion-volume floor, start manual, get the account producing real conversions, then graduate. Putting Smart Bidding on a low-volume account is asking an algorithm to learn from almost nothing.

How does enhanced CPC affect quality score? It does not directly. Quality Score is about ad relevance, expected CTR and landing page experience. Bid strategy and Quality Score are separate levers - do not conflate them.

The gap: every bidding strategy is only as good as the conversions it learns from

This is a Layer 5 problem, and Layer 5 is where the whole stack pays the bill.

Smart Bidding is a machine learning system. You already know that. What gets glossed over is what that actually implies: the model has no independent idea of what a good customer looks like. It knows only what your conversion data tells it. Every conversion you send to Google is a training label that says "this is what success looks like, go find more of it." The algorithm is obedient. It will find more of exactly what you fed it. That is the entire mechanism.

So now ask the uncomfortable question. What are you feeding it?

Two things corrupt that signal, and most accounts have both.

Missing real conversions. Analytics and conversion scripts get blocked. uBlock Origin, Brave shields, privacy browsers, network blockers - they take out a real share of your tracking, with estimates commonly landing around 25 to 35% of conversion signal lost depending on audience. Those are real humans who really converted, and Google never hears about them. The algorithm concludes that traffic source was unproductive and bids it down. You just taught Smart Bidding to avoid some of your best customers.

Fake conversions counted. Of the conversion events that do get collected, honeypot research during agent-traffic surges puts roughly 24 to 31% as bot-originated. Those are not customers. They are automated traffic that tripped your conversion tag. Google records them as wins. The algorithm dutifully says "more of this" and bids up the channels delivering bots.

Put those together and the model is being trained in two wrong directions at once. It is bidding away from real humans it never saw, and bidding toward bots it was told were customers. No bid strategy survives that. Target CPA will optimize confidently toward garbage. Manual CPC will at least not automate the mistake, which is the entire honest case for manual bidding in 2026 - not that it is better, but that it does not amplify a bad signal at machine speed.

Here is the proof moment. A team at PillarlabAI ran a honeypot on a launch waitlist to measure how bad signup contamination really was. 3,000 signups. 77% of them fraud. 650 traced to a single device fingerprint. Now imagine those signups were your Google Ads conversion events - and for plenty of advertisers, signups are exactly the conversion they optimize on. You would be paying Google to find more people like the 77%. The campaign dashboard would show conversions climbing and cost per conversion looking healthy. And every dollar of that "improvement" would be steering Smart Bidding deeper into a bot vein. Garbage in, garbage optimized, garbage out.

The root cause is not your bid strategy. It is structural: conversion data is collected by third-party scripts that mix bots and humans together, with no filtering and no isolation, before that data is shipped straight into Google's optimizer. ECPC, Target CPA, Maximize Conversions - they all drink from the same contaminated well. Changing which strategy you use does not clean the well.

What actually fixes it

You fix the input, not the algorithm.

Collect conversions first-party. When conversion tracking runs from your own infrastructure on your own subdomain, it is far more resilient to the blockers that delete a quarter to a third of your real conversion signal. Google starts hearing about the real humans it was previously bidding away from.

Filter bots before the conversion is counted. Bot traffic gets identified at ingestion and separated out before it ever becomes a conversion event sent to Google. The fake conversions stop entering the training set. The algorithm stops being rewarded for finding bots.

Then - and only then - your bid strategy choice starts to matter again. Feed Smart Bidding clean conversion data and Target CPA can do its job. Feed it filtered, first-party signal and the 30-conversions-in-30-days threshold actually means 30 real conversions, not 20 bots and 10 humans.

That is the shape of what DataCops does: first-party conversion tracking on your own subdomain, bot filtering at ingestion against a 361.8 billion-plus IP database, and clean conversion data sent onward to Google and Meta CAPI. To be straight: DataCops is a newer brand than the legacy analytics incumbents, and its SOC 2 Type II is still in progress, so a regulated buyer should track that. It does not pick your bid strategy for you. It makes sure that whichever strategy you pick is learning from humans.

Decision guide

New campaign or under ~30 conversions a month: manual CPC, get real conversions flowing, then graduate. Low-volume account that has been thrashing on Smart Bidding: drop to manual, you do not have enough signal to automate. You migrated off ECPC and conversions look great: be suspicious - check what share of those conversions are bots before you trust the trend. You run Target CPA or Target ROAS and performance keeps drifting: audit conversion data quality before you touch the target - the model may be learning from junk. Solid volume and you trust your tracking is first-party and bot-filtered: Smart Bidding is genuinely fine, let it run. You cannot say what fraction of your conversions are real humans: fix that before any bid-strategy decision - it outranks the strategy choice entirely.

You are tuning the wheel and ignoring the map

The ECPC deprecation set off a wave of "which bid strategy now" content. Almost none of it asked the only question that decides whether any of those strategies work: is the conversion data real?

Smart Bidding is not magic and it is not malicious. It is obedient. It finds more of whatever you call a conversion. If a quarter of what you call a conversion is a bot, you are paying Google, very efficiently, to scale your bot problem.

So before you spend another afternoon comparing Target CPA to Maximize Conversions: pull last month's conversions and ask, honestly, how many of those were human? If you do not know the answer, you do not have a bidding problem. You have a data problem wearing a bidding problem's clothes.