PRODUCT — SIGNAL
Making podcasts advertisers smarter.
Design Challenge from Knotch
Podcasts have been an interesting area for advertisers recently, as hosts are able to directly communicate their products’ messaging to the podcast’s loyal audience. Design an analytics product for podcast advertisers to better understand their marketing investments.
We’re interested not just in your designs but a little bit about the process leading there, including research, assumptions, or whatever helps you think through the problem. Our products are built for marketers, not data teams, so make it visual & easy to understand. Include as many charts, visualisations and anything else you think is important to help convey the insights of your product.We would like to see something that has high-level polish. Design language should be: contemporary, informative and beautiful.
Most importantly: Have fun! Make it something that you’d enjoy using too :)
I am used to doing in person whiteboard challenges, which get me thinking on my feet, keeping things lean, and always asking for feedback from people in the room. So I was excited to have a chance to take my own personal time to show my design process and create some deliverables at my own pace. I hate when companies ask for it in only 4 hours. How can someone effectively deliver anything that quickly?
For this exercise, I spent approximately 18 hours outside of my day job to work through my condensed design process and it was so much fun.
I’m calling this product Signal.
Researching the industry
I get excited to jump straight into pixels, but if I have learned anything — its that research up front makes design decisions more impactful.
Interpreting the challenge
I read the brief and established a few principles as baseline for the rest of the challenge.
- I will be designing an experience for the persona of podcast advertisers, perhaps a Content Marketer, but not the podcasters themselves or the audience.
- This will be a new product that Knotch will want to release as a part of its marketing tech platforms.
- Information should be as visual as possible, informative, beautiful and relevant to the target persona.
Before jumping into sketching, I wanted to better understand the podcasting world. As an avid podcast listener, I had some assumptions on how others consume content but even with just a little research I realized how profitable this market is for businesses.
Using clever googling to inform
Single Grain posted an article on Podcast Advertising: What You Need to Know, and greelane posted an article on 8 Reasons Podcast Ads should be in your 2018 Marketing Plan. I used these articles to establish my foundational knowledge of how podcasting has evolved, what it means to advertisers, and how to be successful in campaign efforts.
Some fun stats:
- 525,000 Active Podcasts
- 18.5 Million Episodes
- Americans listening to podcasts had doubled
Advertisers expect to spend $500 million on podcast ads in 2020.
More traditional forms of advertising are dwindling in terms of successful conversion rates. The average display click-through rate is 0.06%, which is an example of the banner blindness phenomenon.
Podcasts have become popular mediums to advertise through because of the people listening or watching. This persona embodies the following:
- Deep level of respect for hosts and the content shared
- Highly engaged, and super-supportive
For advertisers, they must do a couple things to ensure their campaign efforts are valuable. (1) Find (or create) podcasts that appeal to their target market, (2) understand the nature of the audience, and (3) tailor the advertisement to suit the desired conversion goals. In the early years of podcasting, most of the people who listened were web developers and technology entrepreneurs. Advertisers that focused on this demographic and knew how to customize their message to their specific needs were the most successful.
Ballpark costs for podcast advertising:
- $18 per 1,000 listens for 15 sec pre-roll
- $25 per 1,000 listens for 60 sec mid-roll
“Know that there is some trial and error in the podcast advertising space. Sometimes you find a podcast with what you believe is a perfect audience with a great host and the campaign doesn’t perform as expected…and then sometimes you are surprised the other way!”
– Dave Hanley, AdvertiseCast
Competitors and Ancillary Products
Apple released a beta platform at the end of 2017 for podcasters to view analytics on the performance of their episodes, and officially announced Podcast Analytics at WWDC 2018. Any podcaster can submit their podcast, and get drillable charts based on aggregate data collected on the Podcasts app.
Midroll connects advertisers to great and reliable podcasters, with a few insights into advertisement efforts and user base. Their competitive edge — strong relationship with reputable Podcasters and Advertisers.
Radio Tail provided advanced metrics and dynamic ad serving technology to ensure advertising in podcasts will reach the right audience. This no longer seems to be active, since the site hasn’t been updated since 2007 and is fairly broken.
Authentic is also trying to help connect brands to shows through personalized recommendations and measurements for targets and objectives.
This proto-persona is only meant to show which assumptions I am looking to challenge first by talking to real customers and prospects. It is important to be as a designer that I also get access to real customers, and not just expect product or support to handle interviews.
Expected capabilities for advertisers
Podcast Advertisers are interested seeing:
- Greater transparency to podcast data
- Listening patterns, including skip rate of listeners
- Potential podcasts that match their target audience
- Demographic and location reach
- Cost of investment, and expected returns
- Tips for expanding reach and better targeting audiences
- Excitement from hosts on products being advertised
Articulating my assumptions
My beliefs about the daily drivers of this prototype:
- Not new to data analytic platforms
- Been involved in initial user research
- Looking for high level visuals and a delightful experience
- Works from their desk primarily, and doesn’t desire to check mobile device for insights on the go.
- Data visualizations are frequently shared with executives who care primarily about the cost of investment.
- Desire for the podcasts they promote with to have greater reach
Assumptions always need to be validated or refuted by observing real end-user behavior. Dogfooding is a great way to feel what customers experience, but there is no replacement for candid feedback and discovery of pain points through user testing.
Technical and product assumptions informing design decisions
- Aggregate data for listeners and podcasts would be available through APIS
- UI choice will be similar enough to existing frameworks
- Bezier curves are easily implemented
- Data from Knowledge can be ported over into valuable insights
Sketching out ideas
Before moving into any sort of design studio tool, I like to put pen to paper to get ideas out quickly and cheaply. Usually I do this for a particular problem, so I reframed the 4 problems in my Proto-Persona in opportunities using the How-Might-We format.
Analytic tools are a dime a dozen, but there are some great leaders in the industry so I looked to them and Dribbble for inspiration on how to visualize data in a meaningful way.
It was finally time to do some drawing, so I ran myself through 4 crazy 8 activities to get my juices flowing. This isn’t something I normally do alone, and always invite engineers and product into for a collaborative approach that could improve delivery and quality. Diversity of thought leads to creative solutions, and usually the best ideas come from the team and not one person.
Finalizing the design
Thinking about the UI/UX Design opportunity posted by Knotch, there were a few things that stuck out to me that I wanted focus on conveying by creating a interactive prototype.
Passion for animating transitions, design chops, and a visual eye
Normally, I design mostly in Sketch but I decided to give InVision Studio a go, because it enabled me to incorporate animations into my prototype. For this reason, I kept the prototype rather flat. Most data analytic platforms let you drill into data sets to seek more specific insights, however, that could be quiet the rabbit hole for a proof-of-concept, and was not my objective for meeting the requirements shared in the challenge.
Additionally, since I wanted to make this feel like a Knotch product — I inspected their marketing website for color swatches and font family.
Checkout the prototype on InVision! (Best experience in Desktop, sorry mobile)
I want to confirm how analysis is done today, by talking more with potential customers to understand how and why that process fails them. Additionally, it is important for the team to understand more about the use case for this dashboard including:
- what they do with this data?
- who consumes it?
- is there a need for whitelabel reports?
- does anyone need to sign-off?
- will someone want to make annotations or comments.
These are all interesting questions to consider investigating further. At this point I recognize that I know very little. However, with a functional prototype, I can take these assumptions and run a usability test with a content marketer from a company like Bombas. Normally, I would pitch the user a task and see how well they can complete it, but this is more of a proof of concept —so I would like to hear more about their initial impressions or desires before running any task based analysis. There isn’t a whole lot of depth to the platform yet, but there is enough to know how to tweak it based on end-user needs.