🎁 Share your data transformation story for a chance to win a prize 🎁 (ended)

Supermetrics team example: 

Previously, our marketing team faced the challenge of tracking region-level ad performance.

Since we include region codes in our ad campaign names, with custom fields we can easily extract those codes, save it as a new custom field and use as a new dimension in the reports.

This allows us to quickly break down performance by region, giving us valuable insights for optimization.

Data blending and custom fields completely changed the game for our agency. While I have plenty of examples to pick from, I’ll showcase the most impactful across our clients and internally: Unified Budget Pacing & Campaign Risk Management:

• Problem: Our agency works with many clients at once, and in our small Ads & Analytics team, we’re tasked with managing dozens of campaigns at the same time, across 6-7 platforms with all sorts of crazy start and end dates. We’ve run into issues with campaigns not spending in full, barely spending, overspending, or not running the correct start/end dates due to input errors. These mistakes cost time, money, stress and frustration. Clients aren’t always patient with these things, whether they’re our fault or not.
   
• Solution: Unified Budget Pacing & Campaign Risk Management, where there’s a single dashboard that shows every campaign in flight, when they start, end, how much they’ve spent, how they’re pacing, and whether they’re expected to deliver in full or not.
   
• Approach:
  1. Religiously consistent naming conventions with key-value pairs, campaign request forms, and Supermetrics. For every flavor of advertising, we have dedicated forms that collect all the information needed to run the campaign. When that form submission hits the system, we have automations that transform the form data into strict naming conventions that we simply copy, paste, and make slight tweaks to ensure they’re correct every time. These naming conventions include the budget of the campaign, the start date and end date, as well as a unique submission ID we can pair up later on.
  2. Thanks to Supermetrics, we ingest as much data as possible from our campaigns into BigQuery, where we use dbt to extract things like the intended budget, unique IDs, start dates and end dates from the naming conventions (you can use Supermetrics’ custom fields for this, too), and pair with data like amount of budget spent, the platform’s campaign/ad set budget, daily budget and the platform’s start and end dates.
  3. Once we’ve boiled down each platform’s respective data to a consistent structure, we blend all those tables together on date, platform and advertiser, to build one big table loaded with all of that lovely pacing data
  4. From that big table, we calculate the special flags that really make the system work: Flight length (end date minus start date, presented as a whole number), Days left (end date minus current date), Yesterday’s spend, Yesterday’s impressions, budget pacing (amount spent divided by budget, presented as a percentage),and flight pacing (days delivered divided by flight length, as a percentage). Through some math and formatting, we also have some columns that look at all of that data and pump out whether a campaign is pacing perfectly, slightly underpacing, slightly overpacing, or is in need of intervention. 
      
• Results: By reviewing this pacing & risk dashboard daily, we catch input errors faster, we can course correct campaigns that are having trouble delivering, and most importantly, we’re not wasting any budget outside of what clients have requested, and we’re fulfilling campaigns exactly as intended, whether Meta’s having a meltdown or not. Being able to forecast campaigns for clients at scale across all platforms is massive, and we can sprinkle some of this tech over to them so they ask us less questions about budget, and more questions about impact.

​@gcfmineo we’re so happy to see you team’s use case example 😍

Especially this 👇🏻 Music to our ears!

we’re not wasting any budget outside of what clients have requested

Thank you for sharing.

​@gcfmineo As usual with amazing insights! Thank you for sharing this with the Community!

Congratulations to our one and only ​@gcfmineo, the winner of the data transformations story challenge! 🏆

​@gcfmineo I’ll send a gift car to you by email shortly 📮

​@Mariia ​@Sahra thanks so much as always, looking forward to sharing more of these in the coming months!