Enterprise software vendors have no taste

Alex Votsmush has it.

Enterprise software vendors have no taste

Every time I research some domain related to enterprise and check various products in it, my eyes bleed. I just can’t embrace unsightly screenshots on product pages and ugly tours. Why is that? We have so many beautifully designed products, web sites and services on the mass market, but if the product is designed for other companies (especially large) it is almost inevitably looks terrible.

Sprint board in Polarion software. Italic, strike through, bullet points — total mess.

My current theory is that enterprise software vendors have no taste. CEO, VP of development, Product managers that focus on enterprise market — all of them have no fucking taste. There is no taste in companies DNA, nobody cares about design and aesthetic. Profits, revenue, sales and new features — yes! Beautiful design — no.

I have a very simple criteria to check how good the solution is: If I can demo it to the potential customer and be proud about what I show — it is good. If I can’t imagine demoing it — it sucks.

Sometimes initial look is not so bad, but when you start to run real cases you feel a constant pain caused by delays, missed UI patterns, unintuitive ways to do things and lack of hints… stop, it is more about user experience and we are talking about taste.

Rational Team Concert has home icon in top corner. Why?

Why so few enterprise vendors praise good design and spend time on polishing it? Is it so hard to add correct spaces between form element and labels? No, it is not, but nobody cares.

DevSuite has classical UI with overwhelming amount of details, gradients, embarrassing lack of spaces in a form and total lack of taste.

You may argue that it is economically unwise to spend much time on design, train people to differentiate good design from bad design, train yourself to spot shit as an executive. I can agree with this argument for B2B markets, but I personally can’t create products that look like they were designed without passion. Every product should be created with passion and it spreads to every little detail in the product. Period.

Rally Software not the worst example, but still many small details uncover the truth — lack of taste.

Is there any opportunity to companies that value good design on B2B markets? Several years ago I’d said no, but with all latest trends I think design is becoming more and more important. Mobile surge and great B2C products slowly change people expectations from all software products. When you use an iPhone or even a Galaxy S and then switch to Salesforce on your laptop—you feel the difference.

Famous Salesforce is not an exception. Totally successful business has outdated and tasteless product. At least I know they are improving in other areas.

This feeling accumulates with time. You start to notice small glitches and overall clutterness. Eventually, when you’ll see a good looking B2B product you’ll want to give it a chance.

Another argument I hear often is about utility. It is easy to burry utility under the beautiful picture. But good design has both:

The most elegant solution will yield a design that is gracefully tempered with restraint and precision — both useful and beautiful.

I hope enterprise vendors that focus on design and put significant effort into it will win eventually. Unfortunately, this strategy doesn’t bring significant additional revenue right now. But money is not everything in this world. I value passion, craftsmanship and quality more.

P.S. Hacker News thread https://news.ycombinator.com/item?id=9190195

1-click vs. 2-clicks design

I was always confused by two buttons in Google Doc editor: highlight text and highlight background. I constantly made the same mistake again and again clicking on the wrong button. I used this feature several years, but still did wrong clicks…

This year they changed icons and I finally learned them!

The second icon is very good, it clearly shows that background color is here. First icon was unchanged (almost) and it is still wrong. So I learned the second icon finally and rarely click to the wrong one.

However, recently Google changed that and combined both actions into a single button. Am I happy? Hell no! Initial design was good, just with wrong icons. It was a 1-click design. New design is a 2-click design and I don’t like this change.

Interestingly, I never hit wrong icon in MS Word. Maybe it they have good icons? Pages also has a clear solution:

While I think Google Doc Editor is the best online editor, it got slightly worse this week.

UX Rule #1: Everything is Possible

Software systems work with data. I believe great system should obey the one simple rule

#1. Every possible operation with data does exist

The knowledge of possibility is a really powerful thing. I want to create a software system that has almost no restrictions in a given domain. You should be able to extract any data, present data and change any data.

Let’s imagine you are using an email client. Suddenly, you want to group all incoming messages by sender. You should be able to do that. Or you want to find all emails for October, create October folder and move the emails into this folder. That should be possible as well.

Every software system has a domain model behind and a set of rules. As a minimum, system should provide all kind of operations on this model: change properties, change relations, create entities, etc. Unfortunately, almost all existing systems are restrictive. 

If you know that operation is possible, you’ll find it. This knowledge gives you a confidence. The operation can be quite advance and well hidden, but still it is much to know that it does exist than dealing with illogical restrictions. Why the hell I can move a user story into a project, but can’t change it’s owner? Why I can change features in batch, but can’t do that with tasks? 

If the rule #1 is applied, you are giving many great things to the end users:

  • Confidence
  • UI consistency
  • Power

Pretty cool it seems…

Hacker News Redesign: v.Next

I read the post about Hacker News re-design and didn’t like it. The resulting design is not good. It might be better than original design for large screens, but it distracts people from information. Orange numbers, large vote buttons, quite poor information density… I am not a designer, but the problem looks quite simple and interesting. 

Here is my attempt to create a better design for Hacker News. I am aware about some weak sides in this solution and will enumerate them in the end, be patient and don’t fight back right away.

First, I asked myself, how people decide that an article is worth they attention? The process is close to this one:

  1. Read the title – is it interesting to me?
  2. Check points – how other people voted?
  3. Check comments – are there any hot discussion?
  4. Check age – is it fresh or not?
  5. Check source – is it credible? (well, this is somewhat stretched)

Currently you have to scan through all numbers of points, comments and age, but that is quite slow. You have to read every number to do a comparison — quite high cognitive load. Is there a better way? I think there is. I took two most important variables and encoded them:

Age is encoded with color. Dark color – fresh news (less than an hour). Light color – old news (half a day or more). Points are encoded with size. Obviously, large circle shows that a post is popular. I took all points and age values from the first HN page and analyzed them. It appeared that 5 groups are enough for age, while 6 groups for points may work.

Another idea was to highlight titles. Here I took a theme of Readable HN Chrome extension and modified it. Each news item is a “one-liner” with a large title, so you can scan items easier. Comments are aligned to speed up comparison as well. Here is the final result:

You can find important and fresh articles really quickly. This design provides a high level view on all news.

I explicitly removed numbers like 1..30 from the page. I don’t think they are useful. Something much simpler can be added to show page number.

Here are detailed explanations:

When you put cursor over circle, you see exact points and age information as well as vote up button. This information is hidden initially to make scanning easier, but is available when you need it.

There are no colors in the design. Colors will attract too much attention. It is possible to highlight very important things in general, but I am not sure what can it be on HN.

Here is HN original design for comparison:

So what are know problems? 

  • Information density is not as good as in original HN design. It can be improved I think without eliminating the original idea.
  • It is not obvious out of the box that you can vote for news. You should put cursor over a title to see the action. But I don’t think that is a problem.
  • Page number should be visible somewhere. 
  • Ability to flag a submission is missed. But it can be easily added on a cursor over near vote up action.

Overall I am happy about the result. Now you can load up your guns, folks.