Findings: what Dot flags automatically for you

Findings are cards from Dot that surface the handful of anomalies and opportunities across your accounts worth your attention. Findings used to be called Sparks. A finding never changes a budget, and nothing is applied unless you press Apply on a fix.

Think of a finding as a note from a senior analyst who watched your accounts overnight. Each one carries a short headline, a one-sentence diagnosis, a specific recommendation, and a colour-coded severity badge (info, success, warning, or critical). Findings read across all eight platforms: Google Ads, Meta Ads, LinkedIn Ads, Microsoft Ads, TikTok Ads, Reddit Ads, Amazon Ads, and X Ads.

Where you'll find them

Findings show up in four places:

The Findings card on your Home page. Click a card to open the finding.
The Findings tab of your inbox, opened from Dot's icon in the top bar. The same panel holds your alerts on its Alerts tab.
The Findings card on an account's or a portfolio's Overview tab, with the three newest.
The Findings tab on each account, with full history and search. A portfolio's Findings tab shows the same table for every account in the portfolio, with an Account column, and you can review fixes from it too.

New findings can also reach you through Slack, email, and the Bell. For a fuller map of where these sit, see understanding the Pace dashboard.

What Dot flags

A finding appears only when something is significant. Roughly 90% of the time the analysis returns nothing at all, which is what keeps the feed worth reading. When one does fire, it's usually one of:

Cost and results anomalies: CPA (cost per acquisition) climbing week over week, ROAS (return on ad spend) sliding, a conversion campaign spending with zero conversions, or conversion volume dropping while spend holds steady.
Positive signals: conversions up while cost holds or improves, or an ad resonating on click-through.
Platform-specific issues: a single Google Ads keyword eating budget with no conversions, Meta Ads creative fatigue, or LinkedIn Ads lead-form drop-off.

Findings inform, and a few can fix

Most findings are a sentence in the card: a recommendation you act on yourself. Findings never move a budget. For eleven cases Dot goes one step further and attaches a ready-made fix:

Add negative keywords to a Google Ads campaign, from search terms that spent with no conversions.
Pause a Google Ads keyword that spends with no conversions.
Pause a Google Ads ad that loses clearly to a sibling in the same ad group, on accounts whose goal is conversions.
Pause a Meta Ads ad that loses clearly to a sibling in the same ad set.
Pause a worn-out Meta Ads ad: shown often, with clicks falling.
Turn off Search Partners on a Google Ads Search campaign when partner sites convert far worse than Google Search over 30 days. Google Search must have 10+ conversions, partners need at least 20 clicks and no more than a quarter of the campaign's search spend, and the account's goal must be conversions. Google Search itself is never touched, and the campaign may spend a little less.
Stop showing ads on a device (mobile, desktop or tablet) on a Google Ads Search campaign, by setting its bid adjustment to -100%. Over 30 days the campaign needs 10+ conversions, the device must be no more than 30% of its spend, and it must have no conversions after spending 3 times the campaign's cost per conversion, or 2+ at twice that cost. Another device must keep converting, and Pace never switches off a second device on the same campaign. Conversion-goal accounts only, and the campaign may spend a little less.
Exclude an age range or gender on a Google Ads Search campaign. The campaign needs 16+ conversions over 30 days, the segment must be no more than 25% of spend (not counting Unknown), with the same cost rules as devices. Unknown is never excluded, Pace won't exclude every age range or both genders, and another segment must keep converting. Conversion-goal accounts only, and the campaign may spend a little less.
Turn off placements on a Meta Ads ad set that spend without converting, such as one Reels or Stories slot. Dot picks up to three individual placements, never a whole platform. The ad set needs 10+ conversions over 30 days, each placement must be no more than 25% of its spend (30% for all of them together), and it must have no conversions after spending twice the ad set's cost per conversion, or convert at twice that cost. The ad set must use manual placements.
Exclude an age band or gender on a Meta Ads ad set. The ad set needs 16+ conversions over 30 days and the segment must be no more than 25% of spend, with the same cost rules as placements. Meta can only trim ages from the youngest or oldest end, so a band in the middle can't be excluded. Unknown and 13 to 17 are never excluded, another segment must keep converting, and it isn't available on Advantage+ audience or special ad category campaigns.
Lower bids at weak times of the week on a manual-bidding Google Ads Search campaign, by 20% to 50%, never stopping ads. The campaign needs 30+ conversions over 30 days. Dot picks up to three windows of at least 3 hours, each costing at least 25% more per conversion than the campaign overall, with no window above 15% of spend (30% together). The deeper the gap, the bigger the cut. Ads keep running at every time of the week.

The two Meta fixes count only your account's goal conversion event, so your Meta account needs a specific conversion event set as its goal for them to appear. Each one restarts Meta's learning for that ad set, and the review panel tells you before you apply.

Fixes only pause things, switch off a network or ad slice, lower bids at weak times, or add negative keywords; they never touch budgets. Budget-changing automation is a separate system, covered in how Pace decides budget changes.

Applying a fix

Nothing is applied automatically. A finding with a fix shows Review: ... in the Fix column of the account's or portfolio's Findings tab, and a Fix ready line with Review fix in your inbox.

Click Review to open the side panel. It shows the evidence and the exact change.
For negative keywords, untick any search term you want to keep, and choose Phrase or Exact match.
Press Apply. Pace re-checks the live data, makes the change, and records it in change history as a Dot change under your name.

Any workspace member with edit access can apply a fix. Press Dismiss if you don't want it. You can add a reason, and Dot stops suggesting that fix.

Undo

Every applied fix shows an Undo button on its row in Changes. Undo re-checks the platform first, then reverses the change: negative keywords are removed, paused keywords and ads are switched back on, Search Partners is switched back on, a device's previous bid adjustment is restored, an age or gender exclusion is removed (even if someone re-added it by hand), Meta placements are switched back on (only the ones Pace turned off), a Meta ad set's age range or genders are put back (Pace won't do this if someone changed them since), and the earlier Google Ads schedule is put back exactly, including no schedule at all. Undoing a Meta placement or audience change restarts Meta's learning again. Undo can briefly raise a campaign's spend until Pace's next pacing run. Undo stays available even if your plan is later downgraded.

Why a fix might be out of date

Pace checks the live account just before it applies. If things moved since Dot wrote the finding, the fix says Out of date and nothing is changed. Common reasons are that the keyword or ad was already paused, Search Partners was already off, the device or age or gender was already not showing, this campaign has stopped showing on another device, the ad set or ad group changed, the placements or the age band or gender were already off or excluded, Meta can't skip an age band in the middle (it can only trim ages from either end), bids were already lowered at those times, the Google Ads campaign no longer uses manual bidding, or the numbers no longer show the problem. If only some search terms have converted since, the fix narrows to the terms that still qualify and shows you the smaller list before you apply.

Why the feed stays calm

Findings avoid repeating themselves. When one flags a problem, you hear about it once, then it goes quiet for three days even if the issue persists. If it's still live after that, Dot re-flags it once so you know it hasn't been resolved. Pace tracks this quiet period per account and per platform, so a noisy issue on one account never silences the same pattern on another. A brand-new critical finding always comes through regardless.

That quiet period is automatic and temporary. Muting is your own decision, and it holds.

Mute a finding you've already decided about

When a finding tells you something you already know and can't act on yet, mute it. Hover any row in the Findings log, or any finding on Home, and open the three-dot menu. The same two actions sit in the footer of a finding's detail dialog as labelled buttons.

Mute findings like this silences future findings about the same kind of issue on that account. Pick For 90 days (the default), For 30 days, For 7 days, or Until I reactivate.
Hide this finding removes only that one row and leaves future findings alone.

A mute is scoped to one account, so the same pattern on another account keeps reporting. Pace also hands the investigation your active mutes before it starts, so it chases a different question instead of rediscovering the one you've settled. A finding that escalates to critical still reaches you: muting at warning or below never holds a critical back.

Muted findings stay on screen. A neutral Muted pill replaces the severity badge and the row dims. Click the pill to see when you muted it, when it expires and the severity floor, then Reactivate to bring it back, every row that mute covered returns to normal at once. Muting and reactivating are both written to your workspace's change history, so teammates can see who silenced what and until when.

Note: Findings are available on the Pro plan and above. On Basic, the Findings card on Home shows an upgrade prompt instead. Dot checks each account once every morning, in that account's own timezone, so findings arrive with your day rather than the instant something changes. There is no schedule to set. Newly connected accounts have Findings switched on by default; accounts you connected earlier keep whatever setting they had.