TDWI | Training & Research | Business Intelligence, Analytics, Big Data, Data Warehousing

5 Minutes with an Analyst: Stacy Caprio of TimePayment Corp.

Analysts must test, test, and retest. Stacy Caprio of TimePayment Corp. and Accelerated Growth Marketing explains the role of an analyst and how she works.

• By James E. Powell
• October 21, 2016

When there's no clear "winner" in your results, you must run even more tests and perhaps go in new and unexpected directions. Stacy Caprio spoke with Upside about her role as an analyst and some of the unexpected results she has found. She is the search marketing manager at TimePayment Corp. and the founder and CEO of Accelerated Growth Marketing.

UPSIDE: What's your favorite part about being an analyst? Your least favorite part?

Stacy Caprio: I love analyzing data and finding a clear pattern or winner. It is always fun to look at the data and see one of your tests win by a statistically significant margin. The even better part is removing the losing test and watching your winning test go on to have a much higher ROI or lower cost per acquisition (CPA) than you were receiving previously. That always makes me smile.

My least favorite part is running a test where there is no clear winner, which actually happens more than I would like. Then you have to either choose a different facet to test or go in a different direction with your current inconclusive test. It is always disappointing to see you've run two variations but neither receives a better click-through or conversion rate. I always want my campaigns to be improving, so receiving no clear direction from a test can be disheartening.

If you could go back in time, what's the one thing you would tell yourself as a new analyst?

You don't have as much common sense as you think you do. Run tests even when you think they are unnecessary because oftentimes your gut instinct will be wrong and it is better to go with data that is correct than an incorrect assumption about ad copy, etc, that you happen to have.

There is one example I would go back in time to test sooner than I did. I used to assume any ad with a © symbol would outperform the equivalent with no copyright symbol so I would add a blanket © to all ads whenever applicable.

When I put this to the test out of sheer curiosity (and partially so I could have data to back up my assumption), I found that is not the case. I tested the © symbol while working with e-commerce clients and found that some products and verticals performed much better with the © symbol and others performed better without.

I found that automatically adding a © or other trust symbol is not the best idea. Instead, it is better to test one ad with a trust symbol and one without for all ad groups in a campaign and then choose the better-performing one by campaign or ad group.

Whether it's the latest Python build or a 50-gallon drum of espresso, what's the one thing you can't do your job without?

I live in Excel. I primarily run and analyze paid ad performance and Excel is where I do all my ad copy, image, time of day, geography, position, multitouch, attribution, and ROI analyses. One thing I currently have no experience with that I would love to learn would be SQL so I could pull data more easily and quickly into reporting.

This is not an issue for any of the standard ad reports including copy, image, time of day, location, position, or standard ROI reporting. However, as I have begun running multitouch attribution reports, the limits of Excel -- including its processing speed and capabilities -- are becoming clear. To perform analyses that involve multiple data sources and multiple touchpoints per contact involves too many data points for Excel to process on its own.

Are you working on anything interesting right now? If not, what's your dream project?

The coolest project I am working on right now is multitouch campaign reporting. Our company just set up Marketo cookie plus URL-parameter paid search application conversion tracking, meaning we can track specific form fill-outs to specific paid search channels.

For example, if John Smith searches for us on Google, clicks one of our Adwords ads, then fills out an application either right then or within 30 days on that same browser without clearing his cookies, he will be identified as an Adwords Marketo touchpoint so we can later link our Adwords applications to revenue.

Our multitouch campaign reporting is an effort to look at all the touches that a person goes through before filling out one of our applications or before contributing revenue to our company.

How it works is we collect multitouch campaign data in both Salesforce and Marketo and sync the two. Then we can run reports in both databases to attribute all touchpoints associated with each applicant. These touchpoints include Adwords, Bing Ads, Facebook ads, retargeting, organic search, sales team outbound calls, outbound collateral sends, and CallRail inbound phone tracking, among others.

We are then able to look at and run analyses on an overview of all touchpoints associated with a single person, which we can use to get a better idea of attribution and how touchpoints affect conversions and revenue.