November 17, 2026
Most PPC accounts do not fail because of a bad keyword or a weak headline. They fail because nobody decided what to touch, when to touch it, and what to leave alone. Good ppc account management is a set of rules that keeps you from reacting to noise and starving the campaigns that are actually working.
This article gives you the framework: what to monitor at what cadence, how much you need to spend before a test means anything, and the failure modes that quietly drain budget. If you read nothing else, the short answer is this: make fewer changes, make them bigger, and give the algorithm enough conversions to learn from before you judge it.
First, a definition, because everyone counts differently. A conversion is the action you decided counts: a purchase, a qualified lead form, a booked demo. Pick one primary conversion per campaign and hold it steady. If you change what counts as a conversion mid-flight, none of your historical numbers mean anything.
What PPC account management actually is
PPC account management is the ongoing work of allocating budget across campaigns, feeding the bidding system clean conversion data, and cutting what loses so the winners get more. It is not daily keyword fiddling. It is portfolio management with a feedback loop.
The job breaks into four recurring tasks:
- Budget allocation.Deciding which campaigns get more and which get cut.
- Signal quality.Making sure the platform is optimizing toward the right conversion, with accurate values.
- Creative and query hygiene.Refreshing fatigued creative, pruning bad search terms, adding negatives.
- Measurement.Reading results against a fixed attribution window so you compare like with like.
Everything else is a subset of these. If a task you are doing does not move one of the four, it is probably busywork.
The cadence: what to check and when
The biggest mistake is checking everything every day and acting on random swings. Daily data is mostly noise. Set a cadence and hold to it.
| Frequency | What you look at | What you are allowed to change |
|---|---|---|
| Daily | Spend pacing, broken tracking, disapprovals, obvious anomalies | Fix outages only. Pause nothing on one day of data |
| Weekly | Search terms, negatives, CPA and ROAS by campaign, creative fatigue | Add negatives, adjust budgets, swap tired creative |
| Monthly | Channel mix, audience performance, bidding strategy fit | Restructure, change bid strategy, reallocate across channels |
The rule that saves accounts: do not change a campaign while it is still learning. Most automated bidding systems need a run of conversions before their output stabilizes. Edit the budget, the target, or the creative mid-learning and you reset the clock. Let it finish, then judge it.
How much you need to spend for a test to mean anything
A test needs enough conversions to separate signal from luck, not enough clicks. Clicks are cheap confidence. Conversions are the currency.
As a working floor, aim for at least 30 conversions per variant before you call a winner, and more if your conversion is high-value and low-volume. If your target CPA is 50 dollars, that is roughly 1, 500 dollars of spend per variant to reach a read you can trust. A two-variant creative test, then, needs on the order of 3, 000 dollars before the result means anything. Run it on 300 dollars and you are reading noise.
Two consequences follow:
- Small accounts should test fewer things, bigger. If you cannot fund a clean test, do not run three at once. Pick the one lever most likely to move the number.
- Low-volume, high-value accounts (common in B2B PPC: Build Campaigns and in regulated verticals like Healthcare PPC: Rules, Costs, and What Converts) may never hit 30 conversions per week. There you optimize on a leading proxy, a qualified-lead event or a micro-conversion, and treat the final sale as a slower lagging check.
Always attach the window to the number. A 7-day-click ROAS of 3.0 and a 28-day-click ROAS of 3.0 are different claims. For considered purchases, a 7-day window undercounts conversions that close later and makes good campaigns look worse than they are. Pick one window per account, write it down, and compare everything against it.
Feeding the bidding system clean signal
Automated bidding is only as good as the conversion data you send it. Garbage signal produces confident, expensive, wrong decisions. This is where most account management effort should go, and where most of it does not.
Three checks, in order of impact:
- Are you counting the right event?If the platform optimizes toward form fills and half your form fills are junk, it will buy you more junk. Optimize toward the event closest to revenue that still has enough volume to learn from.
- Are the values accurate?Value-based bidding needs real values. A flat value on every conversion tells the system every lead is worth the same. It is not.
- Is tracking actually firing?Tracking breaks silently after site changes. A dropped tag can look like a performance collapse. Before you touch bids, confirm the pixel or conversion tag still fires. Coordinate deploys with whoever owns the Web Development so a release does not quietly break attribution.
Common failure modes
These are the patterns that drain spend while the dashboard looks busy.
- Over-management.Daily edits that reset learning. The fix is the cadence table above.
- Chasing last-click.Pausing everything that does not convert on the last click kills the upper-funnel campaigns that fed the winners. Judge against your chosen window, not the last touch.
- Budget spread too thin.Ten campaigns at 20 dollars a day each never gather enough conversions to optimize. Consolidate until each campaign can learn.
- Creative fatigue ignored.Rising frequency and falling click-through on social is fatigue, not a bidding problem. Refresh the creative before you touch the bid. This is routine in Facebook Ads and Instagram Ads accounts.
- No negative keyword discipline.On search, broad and phrase match leak into irrelevant queries. Mine the search terms report weekly and add negatives, or you pay to teach the system your wrong answers.
Different channels fail differently. Search leaks through bad queries; social dies through fatigue; Google Ads rewards signal quality; feed-based campaigns rise and fall on product data. Manage each for how it actually breaks.
When to restructure versus when to wait
Restructure only when the account's shape is fighting its goal, not because results dipped for a week. Signs it is time: campaigns competing against each other for the same queries, budgets that cannot be allocated cleanly because conversions are pooled wrong, or a bid strategy that no longer matches volume.
Before a restructure, look outward too. If a competitor has moved into your top terms and pushed costs up, a sound PPC Competitor Analysis tells you whether the problem is your structure or the auction. And when paid costs climb past what the channel can return, the honest move may be to shift the next dollar elsewhere, which is the whole point of weighing SEO vs PPC: How to choose.
Frequently asked questions
What to do next
Pick one account and write down two things: the conversion you optimize toward and the attribution window you judge by. Then audit the last 30 days of changes against the cadence table and count how many were noise. Most teams are shocked by the number.
If you want a second set of eyes on structure, signal, and where your next dollar should go, start a conversation with the team that runs SilverCrest services.

