Magnetic is programmatic advertising company on Park Avenue in the Flatiron District. Having no experience in advertising technology, I jumped at the chance to learn all about it.
As Director of User Experience, I was brought in to build, from scratch, a world-class customer-facing advertising product. I was the sole designer for most of my tenure, finally able to bring on another designer who had worked for me before.
This is a very data-centric industry. CPA, impressions, click-through, view-through, spend, conversion rate, all these terms were fairly new to me.
While Magnetic had one of the industry-leading databases and an army of data scientists and engineers, they really didn't have much of a customer-facing product.
So if you were a mid-tier advertiser you had to call / email in your requests and get updates.
What they needed was a product that allowed them to get this information on their own.
First Project: Design the MVP
If you're worked in tech long enough you come across this term "MVP" a lot, and it means Minimal Viable Product. It's so widely interpreted as to almost not mean anything anymore. I happen to describe to MLP, which is Minimum Lovable Product, a term I first heard coined by my design buddy Johnathan Shariat of the Design Review podcast. This is the basic product that customers will actually love. You can always add more to it, refine it, etc. But if you launch with anything less than lovable, it might not work.
Head of Product (Keren)
8 weeks (4 two week sprints)
Working closely with my Product Manager and the core product team, we developed a core set of requirements centered around viewing your campaign results.
I drew up workflows and got to understand all the complex data mechanics under the hood.
We knew we needed to deliver something in a few months. Scope creep was a problem but we did a pretty good job at saying "Not now, but later!"
We also needed to design the backbone of our product so we could scale up to allow customers to run their own campaigns. Maybe it wasn't lovable, but it was certainly minimal and viable. Not only that, but most dashboards are awful to look at, and we wanted to differentiate ourselves with some beauty and really delight the millennial audience who comprised most of our customers.
Since I was brand new, I had to earn the trust and respect of my colleagues so I started user testing my early designs and spending a lot of time recording and analyzing them and then presenting them in executive-sized video clips, but always providing source material if anyone wanted to dive deeper.
Some of the questions I asked customers:
1. Describe what you see here.
2. Where do you find your site traffic?
3. How do you change months or look at a particular day?
4. How would you see how your campaign is doing in California?
5. Which devices do your customers seem to use most?
6. Which days of the week does mobile use appear to be highest?
7. Which categories do your customers seem to match up against?
8. What do you love about this product the most? The least?
9. If you had a magic wand changing anything here, what would it be?
10. Would you recommend this site to a friend or colleague?
After processing that feedback and making iterations, we were ready to start building.
This is what we launched with. It's simple, looks great, and does the job of telling our customers what's going on with their campaign and site metrics.
Building Upon Success
After we rolled the product we kept going, adding more features and knocking out our roadmap.
Here, we added in the ability to add money / time to your campaign. This was a sought-after feature that made sense for both us and the customers.
Then, A Curveball
Mergers and acquisitions are often a smart way to grow your company, and Magnetic was no different. So we acquired a company called MyBuys. Software for small and medium e-commerce businesses, they create those "Customers who purchased..." and "Recommended for you..." content areas (amongst other things).
The strategy was to combine their customer data and ours and create an uber company where we could sell programmatic advertising, site recommendations, site purchasing, email recommendations and much more.
So you could be a customer of one, or the other, or both. Quite the product and product design challenge.
So we started a months-long investigation into our new type of customer, and who was hybrid and who was not, and how to separate the product so that it was not just one big mess.
I worked with my designer to create an experience map, which shows the lifecycle of our customers as a result of dozens of interviews with customers and stakeholders.
What's great about these is if you have no idea what's going on, you can see the emotional faces and realize where the pain points are, and attack those.
I quickly realized that no one "product" was ever going to appeal to such a wide variety of people. Some only cared about recommendations, others metrics, and still more were ambivalent.
When faced with such issues, we came up with a solution: personalization. Now, that's a loaded word in many spaces. Personalization is expensive to build. But we did it in an elegant way. Each "card" or "panel" on the Magnetic dashboard could be moved around so if one thing mattered more to you, you could put that on top, or wherever you think that's going to work.
While doing this, we also went under a rebrand, so the colors, typography and logo changed.