A Snapchat Ads audit checks six areas in a fixed order: tracking, account structure, audiences, creative, bidding and reporting. Tracking comes first because every other judgement depends on it. If the Snap Pixel or Conversions API reports the wrong numbers, a campaign that looks weak may be fine and a campaign that looks strong may be wasting money.
The rule we follow is simple: no edits during the audit. Changing settings while you investigate resets learning and makes the before-and-after comparison useless. Write findings down, rank them, then act.
An inherited Snapchat account usually carries history you cannot see from the dashboard. Campaigns may have been paused and restarted, pixels swapped, audiences rebuilt with new names. Editing on day one destroys the clues that explain current performance.
An audit also sets the baseline. When you later report that cost per purchase fell, you need a documented starting point that everyone agreed on. Export the last 90 days of results by campaign and ad set before anything else.
Step 1: Is tracking telling the truth?
Open Events Manager in Snapchat Ads Manager and review every event the Snap Pixel receives. Check volume per event over the last 30 days and compare it with your own records: orders in Shopify, leads in your CRM, sign-ups in your database.
- Does the Purchase event fire once per order, with value and currency?
- Is the Conversions API connected, and are its events deduplicated with the pixel using a shared event ID?
- Do event counts in Snap roughly match backend counts, allowing for attribution differences?
- Are test events or staging domains polluting the data?
- Is the attribution window known and documented, for example 28-day swipe and 1-day view?
A gap of more than a small margin between Snap purchases and real orders deserves investigation before anything else. Use the Snap Pixel Helper extension to check what fires on each page.
Step 2: Does the account structure make sense?
Look at the campaign list with all filters cleared. Count active campaigns, ad sets and ads. Snapchat’s delivery system learns per ad set, so dozens of small ad sets with tiny budgets each struggle to gather enough data.
Structure checks
- Is each campaign tied to a single objective, such as Sales or Leads, not a mix?
- Do ad sets overlap in audience, competing with each other in the same auctions?
- Are budgets set at campaign or ad set level, and is the choice deliberate?
- Are naming conventions consistent enough to read reports without opening each item?
- Are old test campaigns still running on small budgets?
Overlap is common. Snapchat’s audience overlap tool is limited, so compare targeting settings manually. Two ad sets targeting the same age range and interests in the same country with similar creative will raise each other’s costs.
Step 3: Which audiences are actually being reached?
Review each ad set’s targeting: location, age, gender, device, Snap Lifestyle Categories, custom audiences and lookalikes. Then check delivery breakdowns by age and device in reporting. Settings tell you what was intended; breakdowns tell you what happened.
Pay attention to age targeting. Snapchat’s audience skews young in many markets, and an ad set targeting 18 to 54 may deliver mostly to the youngest segment because inventory is cheapest there. If your buyers are older, that cheap reach is not cheap in practice.
Custom audiences need a freshness check. A customer list uploaded a year ago and never refreshed will match fewer people every month. Note the upload date and match rate of each one.
Step 4: Is the creative built for Snapchat?
Snapchat is a full-screen, vertical, sound-on environment for many users. Ads repurposed from horizontal video or static banners usually underperform. Open each active ad in preview and watch it the way a user would.
- Vertical 9:16 format, with key content away from the top and bottom safe zones.
- Brand or product visible in the first two seconds.
- Captions or on-screen text for viewers without sound.
- A clear call to action that matches the attachment, such as Shop Now with a web link.
- Several creative variants per ad set, with at least one refreshed in the last few weeks.
Check frequency alongside creative. High frequency with falling swipe-up rate is the classic sign of creative fatigue.
Step 5: Are bidding and budgets set up to learn?
Snapchat offers auto-bid, target cost and max bid strategies, depending on the goal. Note which each ad set uses and why. Auto-bid needs enough conversions to work; target cost needs a realistic target based on history, not a wish.
Check how often budgets and bids were changed. Large, frequent edits restart learning. In many inherited accounts, performance problems trace back to someone adjusting budgets daily.
An example finding
For example, an apparel brand spending around 8,000 dollars a month might have twelve ad sets on target cost bids set well below their historical cost per purchase. Most would under-deliver, spending a fraction of budget, while two broad ad sets absorbed the rest. The audit finding would be to consolidate into fewer ad sets and reset targets from real data. These numbers illustrate the pattern rather than describe a specific client.
Step 6: Can the reports answer business questions?
Finally, check what the team actually looks at. A report that shows impressions and swipe-ups but not cost per purchase or return on ad spend cannot guide decisions. Compare Snap’s reported results with GA4 sessions tagged with utm_source=snapchat and with backend revenue.
Note who receives reports, how often and in what format. Many problems persist because nobody downstream was asked to act on the numbers.
How do you turn audit findings into a plan?
Rank every finding by impact and effort. Tracking problems come first regardless of effort, because they distort everything else. Structure fixes come next, then creative, then fine-tuning of bids.
- Week 1: fix tracking and deduplication; document the baseline.
- Week 2: consolidate overlapping ad sets; pause dead tests.
- Week 3 to 4: launch new vertical creative; refresh custom audiences.
- Month 2: review bidding with clean data; set targets from real costs.
Make one class of change at a time where possible, so you can tell which one moved the numbers.
Make one class of change at a time where possible, so you can tell which one moved the numbers.