Stop Wrangling Data: The Power of Normalizing Campaign Data
In 2013, IBM reported that 2.5 quintillion bytes of data are created every day. Which sounds cool, but the truth is there isn’t much you can do with that number — our brains simply aren’t set up to ingest one that big.
But that got us thinking: How much data is produced by a modern digital campaign? So, in December 2018, we did a “data audit” on how much data is produced in one month of one campaign.
The campaign we chose produced 3mb per day — around 93mb over the course of the month — and included over 250 unique ad groups across 15 Platforms and 5 different providers.
Let’s put that in spreadsheet terms:
- 14,000 records (rows)
- 50+ unique dimensions (columns)
This raises a similar issue as that 2.5 quintillion number: how do you wrap your head around this amount of information?
The answer most marketers have is data wrangling (which is a real thing with its own wikipedia entry). Essentially, data wrangling is the process of manually gathering data, mapping and sorting it so it’s easier to make sense of it all.
To take care of these duties, we’ve seen agencies and marketing teams use three strategies:
1. Have a dedicated “Data Wrangler” or a whole team of them.
Which is useful, but expensive. Also, it’s a lot of manual work for people whose talents may be better suited to interpreting and acting on the data.
2. Spread data wrangling out among team members.
Which takes them away from the jobs they should be doing and tends to be unorganized and stressful for everyone.
3. Outsource data wrangling.
Which may be easiest but can also be the most expensive. The biggest problem with this approach is that your team is no longer closest to the raw data, a business goal many of our partners share.
These strategies are manual, expensive and often against long-term business goals. The solution is to stop thinking about this process in terms of “data wrangling” and more in terms of “data normalization.”
What is data normalization?
Data normalization is a broad term with many definitions depending on the context in which it’s used. In the context of marketing, specifically in the execution of digital campaigns, data normalization is the process of making all campaign data speak apples-to-apples.
It’s not just gathering data, it’s aligning it all to provide a true view of what’s happening in your campaign. Say you have a video campaign across YouTube, Facebook and Connected TV. Data normalization not only brings all of the data together, it makes sure you compare views accurately.
This solves one of the biggest problems in modern advertising: differing data hierarchies. The way Facebook outputs data is different from how YouTube outputs data which is different from how your DSPs (Demand-Side Platforms) output data.
What Google labels a “Creative” maybe one layer lower than what the DSP labels a “Creative.” And that disparity dominoes all throughout the dataset. Data normalization aligns all these levels so that what is actually a “Creative” is compared to other actual “Creatives” no matter where it shows up in the hierarchy.
It goes a step further by eliminating disparity in measurement caused by proprietary metrics. This issue is a little technical, and we go in-depth in our recent article, so be sure to give that a read if you haven’t yet.
By solving for the fragmentation of how data is reported, data normalization has the potential to be a gamechanger in your digital advertising strategy.
What you get from normalizing data.
Normalized data provides clarity. It gives you insight into how your campaign is actually playing out because it sorts out apples from oranges from screwdrivers. Otherwise, as explained in our previous article, you could easily be comparing an apple with a screwdriver and not realize it.
These more accurate insights allow you to make more informed decisions about not only campaign performance, but larger business strategy. You gain confidence in the data, from the most granular creative comparisons to their ad groups to the campaign as a whole. Everything is where it should be, in the context it needs to be in.
Say an audience segment isn’t responding to your messaging as well as you had hoped but another audience segment can’t get enough. You now have the data-backed confidence to shift both campaign spend and larger business strategy towards that receptive audience.
Data normalization gives you the tools to do something with your advertising and to more aggressively pursue ROAS. You’ll know what’s working, why and what to do about it.
What you need in order to normalize data.
Manually normalized data doesn’t solve all the issues data wrangling raises. It’s still a lot of work, which is expensive and can take time away from other important tasks.
The key is to automate data normalization using software designed from the ground up to do so. LumenAd is such a software and we’d love to show you how it works.
The gist is that LumenAd is made up of four feature sets: Plan, Connect, Monitor and Present. All play a big role in making data normalization work, but the two that really bring it to life are Connect and Monitor.
The Connect feature set brings any and all data together into one place. In LumenAd, you can connect any data source using three functions: API integration, email integration and custom integration. All three will automatically wrangle all your data and normalize it for you.
This sets the baseline for everything that follows. It’s powerful technology but where it all comes together is in the Monitor feature set.
The Monitor feature set takes all the data you’ve connected and creates intuitive data visualizations. It allows you to quickly understand what’s happening and why.
It’s what turns a spreadsheet of 14,000 rows and 50 columns into charts, graphs and tables you can actually use.
After all, it’s not enough to bring all the data together, you need to put it into context. LumenAd automates this entire normalization process so you spend more time acting on data and less time wrangling it.
What you get from automatically normalizing data.
In short, you save a ton of time. We did a study with one of our partners last fall and they estimated that for one campaign, they put an average of 6 hours pulling and normalizing data.
Automating that process eliminates 6 hours of work and it only snowballs from there. Because all the data is already gathered and normalized right here for you, most of the time creating reports is eliminated as well.
Another partner analyzed their reporting time before and after LumenAd. They concluded that LumenAd reduced their time spent reporting by 50%. Over the course of a year, this amount of time regained with LumenAd is gamechanging.
If you’d like to learn more about how LumenAd can revolutionize your digital advertising, schedule a demo today.
At this point, just about every CMO and agency executive is asking themselves what impact the COVID-19 health crisis will have on their respective industries. The reality is we’re most certainly...
Table of Contents 1. Blacklists 2. Whitelists 3. How They Work 4. How to Make Them 5. Download Example 6. At the End of the DayThe use of whitelists and blacklists has become an industry standard in...
Table of Contents AWARENESS ENGAGEMENT CONVERSION Key Performance Indicators, or KPIs, are guideposts for your digital advertising campaigns. They are metrics that are used to quantify progress...