Designers versus data more than ever deserves its place in the pantheon of great conflicts: the Hatfields vs. McCoys, Android vs. iOS, Social Media Marketing vs. Results, Athens vs. Sparta, the Doctor vs. Daleks, Auburn vs. Alabama, and Fox News vs. reality.
We make this out to be some great collision of disciplines when in fact they are not opposites and they can and should work together.
To do so requires the dropping of a lot of false constraints that we place them in. Its all the lies that we tell others and especially the ones we tell ourselves that stop us from doing the very simple tasks needed to really drive performance. Figure out the lies and figure out what really matters and so much noise simply floats away.
Table of contents
![]()
What is data and what is design?
Data is simply a rational way to make rational decisions which are best for the business. Design is about creating different useful experiences, so in concept they should be best friends and work together to simply do what is best for the business. So why then do they constantly fail to do so?
Why are so many designers adamantly opposed to testing and data in a rational sense? Why does so much time get wasted on heuristics, storytelling, and well – nonsense?
We’re all wrong
The truth is that both sides of the debate are guilty of many sins. Analysts often confuse using data to further their opinion for actually improving business. “Creatives” do not want anyone to interrupt their vision of the way things should work. Both sides often fail to stand up to voices from up high and both are often swayed to change their work to match the current prevailing winds, be it from executives or from the industry as a whole.
Both sides tend to be extremely mentally lazy and rely on what feels good or makes others happy instead of diving deeper and forcing rational decision making. Because of this I want to investigate many of the common “lies”, things that are said to defend a point or to push an agenda that have no basis in a rational decision making world, in order to show just how much of the common talk and knowledge about user experiences is just complete BS.
The Many Lies of Design
Here we go.
It’s About Good User Experience
No phrase has less meaning and yet is used to defend so much as doing something in the name of a good user experience. Just for once, I would love to get 5 people in a room and have them individually define this concept with detail and get them to all agree. It is just a cotton candy saying, one full of air and that rots your teeth, and yet no phrase acts as a shield quite like it.
I am not here to say build an awful experience, I am saying that using empty jargon that you feel justifies opinion is a waste of everyone’s time. What is a good user experience? My answer is always one that allows the most people to do the thing that is best for them and for the business. It isn’t one that gets them to look at a specific thing or gets them to look at an interesting picture, it is the one that accomplishes the most for both parties.
This goes back to why a single success metric is so important. It allows you to frame the discussion away from empty jargon or what people think will happen and instead focus solely on the real purpose of the site and of the action you are taking, to generate more revenue. If the “good user experience” that they are arguing about is better for the user, then more of them will do what they need to do to make your site money. If it isn’t, then empty debating just wastes time and finding that out generates more revenue.
This doesn’t mean go out of your way to piss off your users, but it does mean that you need to challenge every part of what you think makes a user experience “good”. The more rules you put in place, the more you are artificially limiting the results that your optimization and organization can drive.
We Need to Provide a Consistent User Experience
From an optimization perspective, this one comes up the most often. You are by definition changing small or large parts of the whole, and the fear and that this will create an inconsistent user experience gets brought up all the time. You need to have a consistent look and feel to your user’s journey… why is this a thing? Why do you care? This is the definition of a rule for the sake of a rule.
Sometimes a consistent user experience matters. Sometimes it is counter-productive. There is nothing easier than testing to figure out the answer. In almost all cases you will be testing the “inconsistent” experience, so if they are really worried about the impact, it is the worst case scenario that you are testing, not the best.
If you find something you can test and see if the same concept applied elsewhere does improve outcomes or is just a waste of time. Arguing about it before the fact is the same as having a heated debate over whose turn it is to clean the bathroom when the entire house is being flooded. It is a debate, but it is very low on the things that matter and you should probably worry more about fixing the leak instead of who gets to do what later on.
You Should/Shouldn’t Copy Competitors Designs
I think by now all have heard the idiom, “Good artists copy; great artists steal.” Too much time gets caught up on focusing on your competitors’ designs and websites, leading to many to have a visceral reaction to the concept of using parts or all of their designs. The truth is that it is not the copying that is the problem, its the competitors part that is where all the failure takes place.
Why does it matter if a design is your competitors or if it is a completely different industry? The only thing that should matter is getting results and figuring out where to apply resources. This means that you need to look everywhere for things that can add to a test by increasing the beta of your options. Look at Google, look at Amazon, look at small lead gen start-ups and eCommerce and ticket platforms and NPOs and everywhere to find concepts that you can leverage or copy directly to including in a range of outcomes.
In the end it doesn’t matter if an idea is new or a direct copy, it doesn’t matter if your arch rival or a company you never heard of came up with what ended up winning. In fact what ended up winning DOES NOT MATTER at all, the only thing that matters is managing the input and acting on the data to maximize outcomes.
This reminds me of a story of when I was consulting. At one point I was working with 5 of the largest travel sites and airlines at the same time. Without knowing it, all 5 wanted to focus solely on the quick selector item on their front doors, what most called their “cannonball”. At some point all 5 of them brought up at least 2 of the others, and months were spent focusing on that one item.
Once they were done, I was able to get them to test other places and to focus far more on discovery instead of pointless internal debate and you know what happened? Not a single one of the sites ended up with a design like the other and not 1 had the front door – let alone that part of the front door – end up being the most influential page on their site. So much time wasted, and so much energy spent debating what their competitors were doing, all for nominal gains.
You Need to Speak to Your User
I for one am all for removing the need to have your users hear voices in their head. That is exactly what I think of every time someone throws this jewel out – be it a designer, a copywriter, or anyone else. You need to create the most efficient and valuable experience, period. If it involves lots of text or creating a dialogue – great. If it involves a single entry bar and two buttons, then also great.
You know what you should do? Test both and other points in between and let the data tell you. And when you get a larger design that works, then you should test to figure out what influence your single behavior and then optimize that.
I would also add the second meaning of that phrase which is how much you put value into user feedback and qualitative data. While it may inform or give you 1 or 2 new variants that you would not have gone with before, the question is how much you can trust their feedback (perhaps not at all) and what the cost/benefit ration is of that exploration (hint: It is awful). You need to generate as many different and high quality ideas as possible, and so if you need to go that route it is not the end of the world, but it is hands down one of the most misleading and one of the least efficient ways to generate variants.
Good Design is Long-Lasting
Not to throw Dieter Rams under the bus, but why is it long lasting? Why does it matter that we come up with something that will be a static thing that we need to stay with for years?
You have to think of every part of the experience as an ever changing amorphous design that allows your users to change and your site to adapt to them. Getting caught up on what worked in the past, or trying to create something that will work 10 years from now is completely pointless. It matters in what it accomplishes, how much cost it was to create it, and how much you can exploit it. This does mean that the longer you can exploit something does weigh in, but not as much as how fast you can adapt and how large the beta is in the concepts that you test.
This is also why adaptive decisioning agents are going to continue to become more and more powerful. Creating a system that automatically can adapt and change based on the current data, and feeding that system with a large array of quality options helps maximize the long term performance while minimizing the value and consistency of any one idea. The larger the array of concepts and the less rules you put on that system, then the better the outcome.
Make a Design for Each [Persona/Market Segment/Target/Geography]
Persona’s and personalization, the bain of my existence. If it helps you come up with a different design that is part of a larger series of testing, then great and you should allow that type of mental model and concept creation. What makes it irrational is when you get stuck on that concept, or when you refuse to discover if your concept works like you think or just maybe works for an entirely different segment of users.
Should you have multiple experiences? Probably, but not always and only when it is far more efficient to do so. What should the division be? I have no clue and I guarantee you do not either. It may be the persona you spent hours building (just kidding, it won’t be) or it may be based on geography, or referrer, or mobile, or time of day, or browser, or user behavior or the weather or any other way of dividing users.
You know what the beauty of testing is? You get to serve all experiences to everyone and then let the data tell you where to go. You can’t lay your hands on what goes where or to whom and you most definitely can not limit yourself to just one type of experience in a rational and high performing environment.
A great goal for a site should be 4 defined user experiences on their site after a year.
This means two rules that change the user experience based on actual results that proved it was worthwhile and consistent to do so. It is hard to find large meaningful exploitable experiences, implement them, maintain them, and then continue to optimize them. They are extremely valuable when done right, but in all other cases it is just a waste of time and energy and almost always results in lower results than just doing nothing.
Good Design Looks Good
It has become a running joke at my current and last job that whatever people like the most or think looks the most appealing is doomed to complete failure during the test. Time and time again the ugly stepchild option wins, yet there are still people caught up on how much they like the visuals of a specific design. Even worse when the one they like actually does win you get comments like “of course it won, it was the only good looking one.”
Good design is about its ability to do work. It is not art, it doesn’t need to create an emotional response, it just needs to do its job. That means that you can’t look at something and know that it is going to work or not with any certainty. The very act of looking at something and making a judgment is subjective, whereas comparative measured data evaluation should be rational. While you can pick up some patterns about things that have a slightly higher chance to work, the reality is if you cut off what you test or what can possibly test then you are by definition limiting the possible outcomes of a test.
You don’t have to like what wins, you just have to like getting results.
What is the truth about Design and Data?
The simple truth is that data is the best thing that has ever happened for designers. Where before they were stuck with fighting for a vision, for doing everything in their power to push their concepts past reviews and executive feedback.
Where they had to design for what impresses people, now they can free themselves from the shackles that have plagued their discipline for so long.
Designers may not like it, it will definitely be uncomfortable, but if they can embrace it then they can stop worrying about vision and feedback and all of the lies mentioned above, as well as just about every other rule of “good design”.
In many ways the real job of the optimizer is to disprove everyone of these rules as often as possible. In order to do that they have to constantly creating a system that enables many inputs and influencing the range of inputs to maximize the chance for that very thing to happen. Optimization is about building a rational decision system and then not allowing any person or bias to knock it off its tracks.
The key of a great designer is to come up with as many different feasible high quality designs and then hand them off to the users. They don’t have to understand all the users or especially what wins, they just have to be open to many different ways of tackling a problem or even about what problems really matter. The better they can feed a system with quality and quantity of ideas and the less they get stuck on what comes out of the system, the better for all concerned and especially their organization.
Conclusion
One of my favorite descriptions about the difference between art and design is that design is art that does work. As an optimizer and as someone trying to enforce data rationality, the key is that I do not care about the art, I care about measuring just how much work it does. It is not good enough that it does work, it is about picking what does the most work. To do that both sides have to be open to going wherever that data tells them and to go against just about everything that is wired into their brains.
Opinions are nice, results are better.
Related Posts
-
Don't design your own website. No, really. It will suck. You might think that since…
-
Every once in a while a big debate comes along in the conversion optimization industry.…
-
136 CMOs were recently asked "What do you see as your biggest opportunity for revenue…
-
Do you need to be a big company and need large volumes of traffic to be data…
Such a militant and misguided post, I don’t know where you work but In all the places I’ve worked in the past design and user experience always strive for the same goal. there is no “fight”. I think the “it’s us against them” culture displayed here is probably the reason you experience your work this way.
No he is preaching AGAINST us vs them.
Quote:
“We make this out to be some great collision of disciplines when in fact they are not opposites and they can and should work together.”
Yap, but it’s overshadowed by the rest of the post, which is pretty clear.
He titled the article “Lies your designer told you”, if Mr. Anderson is preaching against petty bickering he’s doing a pretty poor job of it. I believe his intentions are good and he makes valid points, but they lack context and tact.
I think his goal was to write an controversial article that would get people talking and garner clicks to this site and his own blog. In that regard he’s certainly been successful, so well done. If his goal was to have a healthy discourse on the use of data and design then he’s failed miserably.
Spot on.
Too much cliche controversy on this blog.
It gets tiring and as you say hampers meaningful, authentic discussion.
Also I am talking about optimization (which is far more then simple A/B testing) versus design. UX and Design are two names for basically the same thing. Design can also come from many places beyond just those two titles.
Good on you for bringing it up. I don’t disagree that there are designers in the world who could benefit from your perspective. I just wish you had done a better job of it. Of course you’re not addressing designers, you’re addressing clients; and the way you’re doing it throws designers under the bus. I can’t deny there is truth here but you miss the point nearly as often as you get it right, or you make your point so abrasively that it can only serve to make relationships between clients and practitioners more adversarial.
Frankly, this article won’t effect me, I’m competent enough to handle clients and get results for them. Though I worry for the rest of the industry.
totally agree with Yaron. Don’t know where you’ve worked Mr Anderson, but sounds like you’ve had a poor time of it. Where I work we always strive for the same goals and are perfectly aware of the difference between good aesthetics and good design. Sounds like you’ve only worked with inexperienced web designers from your post. In fact at times you sound like a madman shouting into an empty bucket… so I’m guessing you are being deliberately provocative to spark a debate (?)
I work with a lot of companies, and see this on a regular basis – especially in large enterprises where internal politics play a big role. So while you haven’t experienced that – doesn’t mean the issue is not there.
Not all issues apply to each organization, but this is something I see in my work with clients all the time.
I think you have a very fundamental misunderstanding of UX and what it is we actually do. You set this piece up as an us vs. them when you named the article, “Lies Your Designer Told You (DATA VS DESIGN)”. Instead you should have named it “Things your designer said that you don’t agree with and therefore must be wrong.”
Your designer’s suggestions should be considered a type of data because they are a design professional and have experience in the profession. And while I’m not saying you always have to go with what the designer says without discussion, it’s condescending to call a design recommendation by a design professional an opinion.
Note: Title decision was my editorial decision – to drive more attention (success!)
And – an opinion is an opinion no matter whether it comes from a designer or the CEO.
Opinion is a guess of what might be the best solution. Now we need evidence whether that’s actually true – and change course if needed.
Same here, I’ve never seen or experienced this “fight” you’re mentioning throughout the article, and I believe the designers you’ve worked with were not actual professionals.
I guess you were just trying to make one point: “don’t rely solely on a single designer’s opinion and include data analysis into your designing process”, which I couldn’t agree more with. However this is not a case of X vs Y, but rather something to think about when you assemble your design team and build your product.
I just wanted to point out that “Good artists copy, great artists steal” is not directly a Steve Jobs quote. It’s credited to Picasso though noone is really sure about that (check quoteinvestigator.com/2013/03/06/artists-steal for more research on the subject)
“Good design … doesn’t need to create an emotional response, it just needs to do its job.” I strongly disagree with this statement.
It’s true that today many tools and sources can help us to finetune design.
We have several User Experience testing tools at hand and by analyzing website statistics we can define the need to redesign several aspects of the site or even decide for a full redesign.
But design invokes emotion. This may be difficult to grasp for analytical and rational people, but even changing colour alone can drastically change a buyer’s mindset.
Analysts can define the need for a redesign through data, but it’s the designer who inserts his/ her gut feeling in the design to call on the right emotion. It’s not an us vs them story of analysts vs designers. It’s us AND them because you’ll be bringing together two different kinds of people with different points of view. At times it may clash but I’ve had good experiences nonetheless.
That is exactly my point with the article. The key is that designers need to be open to going beyond the simple emotion they want, and as an optimizer I don’t care what emotion or how you got there, I simply need to get the highest output.
There is nothing inherently contradictory between the two, but people make it out when it ends up being my opinion vs. yours.
Even color can change drastically change a buyer’s mindset? How would you know if it’s changed the buyer’s mindset?
Well, by analyzing the data for instance.
And this is why, as a designer data is one of my best friends.
Successful? Yes. Irresponsible? Probably.
The placement of your reply buttons is less successful. Did you A/B test that? #snarky
Great looks are pushed hard by designers and supported by the project manager because the person paying the bill will easily approve the job. However, this is how we got into the Flash disaster and many “Responsive” will repeat the same mistake.
Without qualitative data, it’s tough to figure out the why behind the quantitative data. Sure, users will say one thing and do another at times but saying the cost/benefit of that exploration is awful – well, that’s just crap. If you aren’t “speaking” to them in a way that resonates with them (and dare I say taps into an emotional element), you’ve lost the plot. The whole creative vs. quant debate is getting old. It seems to me that being a good optimizer isn’t an either/or proposition.
Im with Jen here. I’ve gotten so much value (insights that lead to results) out of qualitative over the years, Im a true believer. It also doesnt need to take that much time, not sure how you Andrew have been doing it.
Yes. Both pieces of the puzzle are incredibly important. This may be what Andrew was trying to convey but I think the message got lost in the writing. Copywriters are good for something…
Its funny, one of the main sparks for this article was a discussion with my copywritter, who I get along great with. I am always talking about no stories, and he is always trying to come up with them. I am all for him or anyone building a mental model and trying to understand the world, but I am also all for doing everything in your power to disprove that world view, especially because by definition when you disprove that world view you have found something that outperformed what you thought would happen.
One of the things people are missing is that some of my best friends across the industry are designers and copywritters. It has nothing to do with the title or the role and everything to do with people getting stuck on emotional responses or acting in a self serving way.
Design is not about look and feel. Design is about accomplishing your site’s goal s and objectives.
Of course we want our clients site to be beautiful but there is also the use case for the ugly landing page. It all depends on your audience.
The ugly long form sales letter still works. It usually doesn’t work so well if you are selling to a sophisticated audience. Even in this case, we are going to split test.
This is why data is still so important to all designers.
It’s funny, I clicked on the article because of the title – I thought “what’s wrong with designers now?” But the basic premise of this article: that data helps inform decisions and things shouldn’t be arbitrary – I agree with; it’s the whole point of good design. It’s not designers vs. data – I want as much data as possible! I want it to perform well. To help a client make money. Sell product. David Ogilvy, a famous adman once said “if it doesn’t sell, it isn’t creative”.
If there is a great conflict, I would say it’s about good design vs. arbitrary beautifulness – but that beautifulness isn’t necessarily championed by designers, but could be demanded by marketers, by executives, by people who want to copy competitors’ websites without analyzing if it’s a worthy idea. By people, who after being told that a particular beautiful thing does not perform well because of x, y and z and here’s the data that proves it, say “I don’t care, I want it anyway”.
For someone who claims to love the specificity and preciseness that data provides, the author only talks about all these concepts at a high level. This may as well be a buzzfeed article for all the lack of depth that is explored about these concepts.
I agree with everything that has been said in the comments before. Since when have design and data been so at odds? The entire foundation of this article is a strawman argument, therefore all of the other arguments in the article collapse because of it.
“Why are so many designers adamantly opposed to testing and data in a rational sense? “
Who are these designers that are so opposed to data? I don’t think there’s a UX job requirement that doesn’t include the ability to either do generative or usability research.
And exactly how are you defining rational decision-making? Is that survey we send out to our users after they use a website? Maybe it’s a net-promoter score. This data will not tell us anything about users except in this moment, do they like or not like our website. This gives us no information about either how we could potentially fix usability issue, address unmet needs, or provide understand the ideal experience the user wants to have when engage with us.
User experience is a difficult word to define because we work inside these all of these nebulous ideas. We’re not discovering gravity with the help of our science tools, we’re talking with highly irrational individuals who will tell us one thing, and do the other. We’re uncovering what’s real in the moment, not what’s true overall, and pulling out the patterns.
The big issue I have with the article is that the data we gather about our users (without understanding any methods that have been used in these “tests”) is dictated by the people who write the test script, create the prototype, conduct the test, and synthesized the results. There are so many nuances involved in actually testing something that’s completely glossed over in this article. We’re not going ever be able to gather this mythical unbojective data because that’s not the type of work we do.
Data is not a panacea that can tell us everything about our users. At some point we have to depend on our expertise as designers to tell us what’s the best approach. Designers shouldn’t be dictators of design, nor should they be slaves to business and product management. We should be more service oriented, delivering what each needs in a way we know is right.
As a ux designer, I believe in research and data. I believe it’s important to understand your users, their workflow, their frustrations and pain points, and what they’d like their ideal experience to be. When it comes time to actually put pen to paper (or mouse to adobe product) it’s my job to decide how I’m going to be able to make these ideas a reality. But in the in event that I can’t get all of this information, I can still make good design decisions based on my knowledge of the space and my experience as a designer. Data can be a crutch, and only naïve designers can’t work without it.
Let me make it really simple.
If people like your site = Does not matter
Your net promoter score = doesn’t matter
99% of the data people want or usually look at = doesn’t matter
It comes down to very simply RPV. Do not look at anything else, do not pass go, do not collect your paycheck if you are focused on other stuff.
To me the most key line in the entire article is this:
“Both sides tend to be extremely mentally lazy and rely on what feels good or makes others happy instead of diving deeper and forcing rational decision making. ”
I am just as tired or even more so of bad analysts and bad optimizers who make people think they are accomplishing something when they are just spinning wheels. UX/Designers can be good, ithey are usually awful at their jobs. Analysts/Optimizers can be good at their job, most are awful at their job. Even worse is when people confuse the time they spent with the bad apples of one or the other discipline and then think that in someway speaks to the actual value of the discipline.
What Anna said above in a few paragraphs is far far more insightful than any of than any of the dreck in this article.
Somewhat of a rubbish article. Sounds like the author has his own beef with the design, rather than understanding the design has its own purpose and finding the ways to work with it, rather than creating some bizarre data vs. design theory.
You might have read it as a us vs them story, but it’s actually “we’re all in this together” story.
About User Experience. Think about Walmart.
60% of Walmart clients are going there to buy milk and bread. So if you think from UX point then they should place Milk and Bread right near the Entrance (between entrance and checkout zone). So why the hell they put Milk and Bread at the very very far corner of the shop?
Probably because they concentrate on Shopping Behavior and not on UX. Client come to Walmart to save money on milk and bread. boom . thats it.
Yes, it is about RPV, the ability to monatize that person coming in. That being said you could also see what happens when you put the milk at the front, or the back, or the middle, or the far outside of the store, or a few other arrangements and see what has the biggest impact to the bottom line.
IT IS NEVER ABOUT ONE IDEA VS. ANOTHER.
Incredible content about conversions and the awfull too much web designer. l will use as a source on my greek -sorry blog about web design and SEO.
Thanx
I’m so sick of posts about how data is basically the panacea for bad design.
Yes, good designers value data. But they also understand the company holistically. What’s the brand, the vision? Compromising those things for conversions or sales on your website might lead to short-term gains, but long-term losses.
Yet many an ‘optimization expert’ would encourage those changes because it improves the ‘single success metric’ (sales/conversions).
Good business people (and I consider good designers businesspeople) understand the need for both data and more qualitative things like branding, vision, etc.
Data is not the panacea for bad design, and one of the key points of the article is that very fact. That being said hiding behind a hate for data to defend bad design is very counter productive. Using any sort of data to validate an opinion is awful and irrational, and as such should be shot down. The problem is when people hide behind their discipline, be it design, UX, analytics, marketing, etc… instead of admitting they know much less then they really do.
Also as a note, I have yet to meet someone that doesn’t think of themselves as a good business person. It is irrelevant to the actual impact of their work.
The truth is always that we are the heroes of our own stories.
I’ve been on the Internet, selling since 1996 – now in that time I have learned two things (well of the many things I have learned 2 are relevant here)
1) What YOU think matters not a jot, it’s what the customer thinks that matters
2) When you make a “design decision” or say “I think blah…” STOP, refer to point 1.
Statements like “Compromising those things for conversions or sales on your website might lead to short-term gains, but long-term losses.”
are a little short sighted, I have yet to see a customer who made short term improvements to their profits make long term losses. That’s like saying working out and eating healthy now might make you sick in the end.
Do we see amazing, ebay or other large organisations making huge changes? No, because they realise small incremental changes help them to understand their customer and improve sales over the longer term.
Having worked with literally hundreds of websites owners, businesses and designers I can say in my experience designers are solely concerned with aesthetic and do not see the commercial imperative.
I am sure there are dozens of examples that can be given of good “designers” but I’ve yet to personally meet a “website designer” who starts by looking at the Market, creating ideal buyer profiles, creating marketing collateral around their needs and the unique value proposition the business represents and then making sure that user benefits are paramount. Most of these concepts are poorly understood, and have ever been so.
I fear that there are a lot of designers who have commented negatively because there’s a grain of truth within this article. The simple test is whether you use Analytics and CRO to inform the process and make changes based on metrics, measurement and user input rather than just what you “feel” looks good.
I love your stuff, and try and read as much as you post. But really? That first sentence in the 6th paragraph with FOUR inline links? And the sentence is only 14 words long? All links going to different places? ARG. Kill me now. Sorry – it’s my bugaboo.
It doesnt take you anywhere if you dont click :) Its just a color change, i like it actually :)
Great article. This has been my philosophy over the last few years and it certainly pays off. It is great to work with designers that understand this concept as well as it creates the best of both worlds for everyone.
How long do you run an experiment for? Do you have any suggestions? Would be great to know.
When to stop A/B tests: https://cxl.com/stopping-ab-tests-how-many-conversions-do-i-need/
Thank you for the reference Peep
You need to write another article about why people are so hot and bothered about this. Interesting. Congrats for generating such thoughtful (and often derogatory and mean-spirited) comments. But I learned good stuff from these comments, too.
Honestly, I gotta just throw this out there… I’m with the author on this one. Some designers can be “feely” — I’m more of the “let me see the data” type & we have indeed clashed over the years… All in good fun though [mic drop]
Mr. Alexander has taken the FOX News approach to increase the drama, making Mr. Laja very happy, yet leaving readers disappointed. I bet you would rather have impressed us with your knowledge. I hope there is better to come in the future.
Debbie, I am not sure I am following you. Drama? How are you disappointed from this free article?
I’m also curious how is getting evidence / data to back up a designers creative idea, so designers would have more power (the central theme of the article) – “drama”?
Design opposing Data based optimization is like Chicken opposing Egg. Not sure that makes sense… You need to create then measure then create based on the learnings then measure, then create and so on. Or was it measure first?
The article says design shouldn’t be l’art pour l’art but an other tool in achieving your (single) goal. And data tells you if you are achieving it or not.
That is exactly what every design school teaches though.
People pushing their stuff is not exclusive to designers.
One of the key points here is that Design as a whole doesn’t oppose data, but many designer do. If you the thought process is “I am going to create this one thing and then see if it achieves the one goal” then you are fundamentally missing the point (and hurting your company).
The entire point is that there are hundreds of ways to tackle any problem and the role of the designer is to come up with as MANY feasible high quality ones as functionally possible. It is never about one idea, that is just idea validation.
So it is not chicken opposing egg, it is farmer getting a large number of chickens to get the highest yield on number of eggs and choosing the highest quality ones, instead of just relying on a single chicken and a single egg.
So let’s say you have a startup and need a website, you start with 5 different websites, then check the data and go with the best performing one?
Or how do you gather data on which one is the quality one?
There is a functional minimum viable amount of data to act on data. If you are a true start-up then they are all flawed and you are just using your opinion, do one, feel good about it, and accept the fact that it is very low efficiency.
Once you get to the point where you are getting at the very minimum 500 conversions per month, then you would take your most trafficked page and test out 5 as different from possible versions of that page.
Agree, so what we are talking about is data based optimization, where data rules of course, and opinions in general (not only designers opinions) are irrelevant.
On the other hand when you come out with a new product, you need to rely more on creatives, and “expertise” and most likely you will try first with a “good, consistent UX” or at least your concept of it.
When you create, you need creatives, when you optimize, you need experiments and data, and you need to be highly aware in which stage you are.
It certainly requires a specific mindset from the designer.
I enjoyed reading the comments while the article was not a good read and should be deleted from conversionxl in my opinion.
Well done for creating a debate and getting lots of UGC :)
Hi Peep,
Thank you for raising some important, but often overlooked issues. I find that all too often people come up with reasons not to make changes rather than being open to testing whether or not it goes against some brand guideline or other policy that has been invented without evidence to support it.
I wrote a post on brand guidelines which is based upon personal experience. Let me know what you think. http://conversion-uplift.co.uk/post/112779964127/why-are-your-brand-guidelines-costing-you-millions
Cheers
Neal
Having a huge laugh at the comments as my reading of the article is completely different from most people’s (or is it just designers?), apparently.
What the article is saying is essentially this: design is one mental model among others. Sometimes it might yield good results, and sometimes it imposes constraints that might *prevent* good results, as would strict adherence to a set of principles that are not inherently tied to optimisation.
And because design is hip these days, people get butthurt when they face the fact that the end users of their beautifully designed apps may not be so receptive to their art, or not as much as they could be to other factors that are less glamorous.
All in all, Andrew’s points seem to me to be rooted in humility more than anything else: there’s no optimisation silver bullet, and that stands for design too. Which doesn’t mean that “design doesn’t matter” or any such extremist nonsense. Take a chill pill guys. ;)
Hi Andrew,
Interesting article – I would agree with you on most of the points you raise.
However, I think I have to disagree with you where you say “UX and Design are two names for basically the same thing” because although design is obviously a big part of a website, the development and build should be considered equally important.
Obviously the with the article being titled “Lies Your Designer Told You” I wouldn’t expect you to cover much on development, but there are factors which can affect UX that don’t rely on design. Page load speed is a great example – if you have poor hosting or your files aren’t compressed so that the page loads quickly, it doesn’t matter how good your design is the user will eventually leave your website whether you like it or not. This goes in hand in hand with “Good Design is Everlasting” – if you have a poor markup quality and a lack of functionality it can prove to be a real headache when redesigning the website.
Thanks for making the time to write the article Andrew, you have certainly raised some great points for debate. I’ll be sure look out for your other posts in the future.
Interesting article. I do agree with Alex Brooke that UX is equally important to the design.
Great article for designer. Thanks you :)