Back To Blog
Design Sensibility + Comps

On the audio commentary track to "All the President's Men" [link] -- Robert Redford recounts a conversation he had with Bob Woodward when he was researching the film. Redford to Woodward: "'One thing seems to be sure, you're a workaholic. Why do you work so hard?' He said that when he was a student at Yale, had comprehensive tests two days running. The first day's test he hadn't studied for and he blew it and he was really upset with himself. He knew he had failed miserably. The second day's test he studied for intensely and he nailed it. The results came in and he had gotten a good grade the first day and failed the second day. He was in shock and went to the professor and said, 'You gotta be wrong. Show me.' And he did. And from that point, he realized, 'I did not know what good work was.'" (paraphrasing; starts at roughly 17:24 on the DVD)

In college, one of my professors was on the board of Reader's Digest and he arranged an internship there for me. This was 1999 and they astutely realized (or I should say, their CEO, Tom Ryder realized) that the Internet was coming and it would be transformative. They had a small new media group that did the things you would expect a new media group would do for an old line publication company -- digitizing existing properties, etc. -- but they also had a new initiative called They had acquired a number of valuable URLs and wanted to create businesses around them. (I know, it's odd that Reader's Digest would create a subsidiary around giving gifts -- but I'll leave that alone in this post.) I spent most of my summer working on

Reader's Digest had engaged a branding firm to help with creating a logo for In their pitch meeting, this firm showcased the work they did with a major auto manufacturer. They had a few PowerPoint slides which showed dozens of logos. It detailed their thoughts and ideas and how it evolved over time as they refined their thinking and the client gave input. This eventually coalesced into the logo that was chosen -- which was quite good.

One of the things Steve Jobs talked about was how working on product was partially a function of bringing all your other sensibilities -- all the things that you learned from interacting with the world -- bringing that back to your work. In his Stanford commencement speech [link], he talked about dropping in on a class at Reed that dealt with typography -- and how this ultimately informed not only his desire to offer many fonts in the Mac, but also the importance he placed on having the right fonts and beautiful fonts and the work they did. I think that's a huge, continuous part of being a good [fill in the blank in terms of jobs] -- basically anything associated with product. That your own sensibility keeps getting better, more refined, and that you have a larger and larger library of things to draw upon. That you have good taste, and likely, strong opinions about what is good and not so good.

But before that (or even after that) -- how is it that one makes good design choices? How does one know what good work is? This is where I want to stress the idea of comps -- the lesson I learned when I saw that ad agency pitch. If you showed me one logo -- maybe I think it's good, maybe I think it's bad. More problematically, let's say I have a sense of how I felt -- I'm leaning one direction or another. Unless I was a designer, it's unlikely I'll know exactly (especially with an early stage design) -- what I want changed. Is it too big? Too bright? Too sharp? Too complex? Too difficult to see? I'm not sure. I'll likely say something to the effect of, "I like it but don't love it." That's borderline useless feedback.

Let's say, however, that I saw 3 design. Or 5 designs. I am then going to start formulating an opinion about which ones I like and why. Maybe some are too boxy. Or some don't evoke the core concept of the product. Or the color palette isn't quite correct for some reason. I'll then say things like, "I like X from design A and Y from design B." I'll start to be able to have an opinion and give feedback in a way that a designer can use and implement. We'll be able to iterate. 

I was talking with a startup founder earlier today and he mentioned that he was thinking about hiring a UX designer. One of the things that I've found surprising in recent years is how design has evolved. My sense of design is very simplistic. As a product manager, I worked off of two sets of documents. The product requirements document (PRD) -- which basically summarized what the product was and why we were building it, and then a tiered list of features. Tier I was everything we needed to get done in the immediate release. Tier II was stuff that we should get to if we had time. Tier III was stuff we probably weren't going to get to but it was good to have important features listed for reference. The other "document" was mock-ups. These were not wireframes. These were not some sort of intermediate set of drawings. They were non-functioning Photoshop files / jpegs / etc. -- that essentially showed exactly how the product should look (and work.) These were very clear. If we launched a product that didn't look and work like the mock-ups -- there was no confusion. One could easily pull up the mocks and compare it to the live site and ask, "Is this the same? How is this different?"

So why is this important and how has the world changed? I often see wireframes or UX designs or some sort of intermediate mock-up. These roughly show what elements are on a page without explicitly laying out how they will look and function. That's kind of useless to me -- or at the minimum, doesn't add additional value to me beyond what I previously had. The PRD says, "Here's everything we're building and what it needs to do." The mockups say, "This is how it look and works." So where do wireframes fit in? I don't think they do. What additional value do wireframes have over that world? Worse -- wireframes give the impression of shared understanding. You look at the wireframes and go, "Yeah. I think that's what we need. A search box. A place to log in. A section ot merchandise this item. etc." People think they're on the same page. However, give those constraints and you can easily have 1000x different takes on it. Details matter. What ends up happening is some variant of the wireframe appears. The product is built. The PM is unhappy -- it's hard to iterate from that space. You need a space to iterate and a working product is not that space. It cost a lot of engineering time to get there. Battling it out over the designs and working through those issues when things are just in Photoshop is the ideal place.

Getting good looking products is hard -- frankly, you need to start with a great designer. But even with a great designer -- it takes this type of work. It takes a shared understanding of what the product is and how it should work -- but then it takes that process. The way I like to work is that I'll literally sketch -- maybe on a piece of paper or on a white board -- different major screens of a product. Just draw out boxes of what should be on each screen. Basic functionality. I'll then give that to the designer and talk with him/her until they felt comfortable with it. They then create their conception of the product. If you have the resources -- have them come up with multiple versions of the product, or at least the home screen. I'll often tell a startup that what they've sent me is too safe. That their stuff is functional and logical but not inspired. That's what you want -- inspired. Have the designer go to town. Have them work on something without constraints. You don't have to go with it -- but that'll become an option. Or elements of it may be things you like and later incorporate. Have 3-5 versions of the product. A few will likely be close variants. But maybe 1 or 2 are totally out there and it'll get you thinking, "Oh, maybe this is it." Remember Gilt and RueLaLa and their siblings? Those aren't Silicon Valley designs -- they're NYC designs. Giant pictures? Emotional / evocative? Those people come from a different world than the people who designed AdWords. So push those boundaries or at least allow your designer to present you with options. Come from a world where options are available so then you can actively say no -- because if you've never seen it, you won't be able to know if it (or elements of it) might be right for your product.