Before I joined Adverity, I was one of our users. I worked at an agency reporting on several clients at once, and Adverity is how I automated the weekly and monthly reports that went back to each client.
Every promo period meant gearing up for a spike in activity across all those accounts. The biggest lesson I took from it was that a little prep ahead of time really pays off. Promo periods like Black Friday run much smoother when your team is in the loop on changes, and any issues get flagged the moment they happen.
Now, as Customer Marketing Manager, I run our Customer Office Hours and talk to the teams using Adverity every day. Every season, the same questions come up about getting ready for Black Friday. Most teams already have what they need in their setup. What makes the difference is a handful of settings and checks that will make your life easier if you can set them up before the week of the promo.
So what's worth doing now? Here are four common mistakes I see, and how to avoid them.
Holiday campaigns often add sources and markets you haven't run before, which is exactly where harmonization gaps tend to show up. Harmonization means mapping data from different platforms onto the same names, groupings, and definitions, so everything adds up in one report. New sources and campaigns need to be part of that mapping too, and when agencies or new team members help with setup, it's easy to slip up.
Oftentimes, catching a small mismatch early saves cleanup during day-of reporting after Black Friday.
Before launch, run through these questions:
- New data sources: Are they mapped to your existing channel groupings, metrics and KPI definitions?
- New campaigns: Do naming conventions and UTMs roll up correctly in reports?
- New markets: Are currency conversion transformations set up, and do they cover every currency you’re running in?
And if you find yourself making last-minute campaign changes ahead of Black Friday (or any promotional period really!), don’t forget to communicate that to your stakeholders.
I’ve seen so many instances where client teams make changes without informing their main stakeholder or their agency contact. Then, when budgets inevitably underspend or performance numbers don’t add up, it becomes a mid-campaign goose chase to uncover the cause. Overcommunication is your friend!
Over a busy weekend, lots of people make changes at once. If you have a lot of clients or teams working in your datastreams, small mistakes can stop data from flowing, for example if someone resets a password, or edits a transformation for one instance that affects another. Notifications tell you straight away when something needs attention, so you can sort it before it reaches a report.
The notifications teams subscribe to most often include:
- Datastream and transformation alerts that tell you when an issue stops data being collected.
- Authorization alerts which let you know when an expired login causes a fetch to fail.
- Import alerts which flag errors or duplicates when data loads into your warehouse or file storage
- Performance alerts that show when tasks take longer than usual
Pro tip: Over the promo weekend, it's worth temporarily adding a few extra people to your Slack or Teams notification channels. That way an alert reaches someone who can act on it, even if the usual owner is offline.
Most Black Friday purchases happen over the long weekend, but the conversions credited to your campaigns keep arriving for days or weeks afterward. Ad platforms update their numbers as late conversions come in, up to the end of each platform's attribution window.
A daily fetch of yesterday's data won't go back for those updates. If that's your only schedule, your final campaign report can understate what Black Friday delivered, even though the numbers look complete.
This is where Smart Schedule helps. You can now run more than one fetch schedule in a single datastream and choose when each one starts. A setup that works well for peak season:
- A daily fetch that pulls yesterday's data, timed to finish before your team's morning check-in.
- A weekly fetch that re-pulls the last 30 days, so late conversions show up in your reports without anyone chasing them.
Set that lookback to match the longest attribution window you're working with, and keep the weekly fetch running for a few weeks after the promo ends.
Ad platform data tells you what you spent and what it drove, but not always why. TikTok Shop has made some really great enhancements to its APIs, like best seller and creator performance data, which help paint a more detailed picture of performance. I’m also a big fan of sources like Threads, Pinterest, and Reddit for trend discovery and brand sentiment analysis before and after the holiday.
Retail media, marketplace sales, and social listening help explain why a campaign did or didn't convert. If a product sold out on Friday morning, you'll want that context next to your ad spend when you review ROAS.
Sales, inventory, and social listening data add valuable context to Black Friday campaign performance.
Adding a connector is quick, and it pays to start in October rather than the week before, so you can test and fine-tune dashboards before Black Friday.
Some of our new and notable connectors worth a look for holiday campaigns:
|
Category |
New connectors |
|
Retail media and marketplaces |
Amazon Ads, Criteo Retail Media, DoorDash Ads, eBay, TikTok Shop, Roundel, Walmart |
|
Social listening |
Brandwatch, Infegy, Meltwater, Quid, Sprinklr, Sprout Social |
In the last few days before Black Friday, there's always a list of small jobs to get through. Connect MCP lets you hand some of them to an AI assistant like Claude or Gemini. MCP (Model Context Protocol) is an open standard that lets AI tools connect to other software, so you can ask for things in plain language instead of clicking through each screen.
Here's where it helps most in the run-up to the promo:
- Run a pre-weekend check on which datastreams or authorizations need attention, without opening each one.
- Map new data sources or update several datastreams at once with a plain-language request.
- Pull together performance summaries by channel, persona, or any other segment when stakeholders ask how things are going.
See how Elcano uses Connect MCP for agent activation and dynamic campaign optimization.
You don't need to tackle all four this week. Start with the one you're least sure your setup already covers. For a lot of teams, that's late conversions, because a report that's missing them still looks finished.
If you’re confident in the setup and scalability of your data foundation, my recommendation is to focus on MCP or other tools that can take your team’s ability to create and analyze campaigns to the next level.
Or even better, build data apps with Adverity Atlas that transform the way your teams optimize and analyze campaigns. For example, minimum ROAS profitability calculators, or inventory-aware spend checks to ensure you don't pay for clicks you can't fulfill.
The most common are skipping mapping checks for new campaigns and markets, waiting to spot issues on the dashboard instead of setting up notifications, fetching only the latest day’s data, and leaving new connectors until the last minute.
Before launch, check that new data sources are mapped to your existing channel groupings, metrics, and KPI definitions, that campaign names and UTMs follow your naming conventions, and that currency conversion covers every market you’re running in. It also helps to tell everyone who reports on the campaigns about any last-minute changes.
Set up pipeline notifications for datastreams, authorizations, imports, and task performance. Route them to a Slack or Teams channel your team watches during the promo period, and add extra people temporarily so someone is always available to act on an alert.
Match it to the longest attribution window across the ad platforms you use. Many teams run a weekly fetch that re-pulls the last 30 days, and keep it running for a few weeks after the promo ends so late conversions are captured.
As early as you can. October is ideal, because it leaves time to build and test dashboards on the new data before traffic peaks.
With Connect MCP, you can ask an AI assistant like Claude or Gemini to check which datastreams or authorizations need attention, update several datastreams at once, or summarize performance by channel, all in plain language.