3 ms·

You didn't insinuate it, but you also aren't being specific about the kind of help you're asking for. You're making enormously broad statements about at least

You didn't insinuate it, but you also aren't being specific about the kind of help you're asking for. You're making enormously broad statements about at least two different professions that solve different problems in different ways. I don't believe I am being unfairly critical of your phrasing. I'm guessing you're reading into my statement, "I'm guessing when you ask for help with your code, you're not expecting someone to develop an entire application for you, for free." I'm not suggesting you are personally asking for an entire design for free (although you would not be the first if you were). Rather, I'm saying that, if you don't understand your design problem well enough to ask a narrow-enough question of the right sort of designer (as you might with an engineer about a technical issue), then the person you're asking has to understand -- and potentially educate you about -- a much larger portion of your project. That sort of time investment only comes with payment. I was a developer once, a long time ago, and I don't recall ever asking strangers or even friends to review my code. At least, not successfully. Are there actually places you can go and have random programmers look at your code? How do they run it? How do they build it? How do they understand your business logic? How are they able to duplicate your production environments? I feel as if your counterexample is a straw man. Code reviews amongst other people on your production team is not the same thing at all. Dribbble is not a place for critique. It is a place for praise, specifically, between visual designers and artists. Only some of that overlaps with UI. Most people don't post in-progress work there, and Dribbble is not designed to facilitate critique of in-progress work when they do. So, that's not a great example. Also, it's rare to actually be seeking "open-ended design help." If you have an MVP already, then you're already talking with a designer too late. There is a limited amount of change you will be willing to tolerate to your MVP, in very particular places, dealt with in very particular ways. It's the button color example from my first post again. Your MVP might not need a button color change. You might need to throw your whole MVP away. You are probably not prepared to do that. I have talked with perhaps a hundred different startups through office hours in the past three years or so. Most startups want a pat on the head. They're not prepared to make any design-related changes at all, not to colors, not to copy, not to anything. Some startups are prepared to make superficial, cosmetic changes, but most designers aren't interested in helping you put lipstick on a pig. Designers do that all day long as their day job; they want a bigger impact if they're doing it for fun. Nearly no startups are interested in understanding how to figure out the real design problems in their product and how to solve them; that is, how to find product-market fit, and are willing to throw away all their existing work if necessary to do so. I could probably count those startups on one hand. I have absolutely had people get up and walk out mid-sentence when I pointed out irredeemable problems with their idea, MVP, or launched application. That is not the kind of feedback most people are willing to hear "for free." And if you are prepared to throw your MVP away, then you're stuck, because if you ask anything else of the designer at that point, that's real work, and they're not going to do that for free. I think good free help can be obtained if you can ask a narrow-enough question of the right sort of designer, absolutely. In my experience, though, that question is hardly ever the right question you should be asking.

sandGorgon 12y ago
I see. Thanks for taking the time to write this out - and the answer to your question is yes: there are many places for people to look and review your code (not limited to http://codereview.stackexchange.com/ http://codereview.stackexchange.com/ and IRC). From all your answer(s), I get the feeling that UI/UX designers tend not to spend time for free because of one of three reasons: 1. a quick design help cannot be done and the alternative is too much effort to be justified 2. if you have an MVP before you are taking design help - its already too late 3. You really dont want design help - you want platitudes Let's say, for argument's sake, that I dont fall in any of these categories - built a MVP, took it to market, started generating revenue already. But I think I understand design (but am unskilled at it) and I can take a complete trashing of my current work. However, for certain startup-survival reasons, I need to make something incrementally better (both UI and UX). It's not the promised land, but just one teeny step towards it. How can I frame my request for help to not piss you off, not waste too much of your time and still not kill my startup ?
- vitovito 12y ago
  Your second-to-last paragraph is a good framing, and your final question is a good question. It's also a good example of why quick design help is hard, because there is no way for anyone but you to know what would be incrementally better or not for your users. All an outside designer can do is suggest things that might be interesting to test. You still have to try them and measure the change over time. A given designer has no specific knowledge about your market, and something can be well-designed in general but still have a lot of potential for market optimization. ( In fact, a colleague runs a service that is just that: he'll try something new for you, every month. https://draft.nu/revise/ https://draft.nu/revise/ ) Since only you have the data on your MVP, start there. Put together your customer lists. Make sure you understand their demographics and their usage patterns and why they purchased and what they were considering instead. Make sure you have ways to contact them for research interviews and if it's possible to go on-site to see them use your product live. Document how your competitors present and discuss their products. Then, pick something. One small thing. A landing page. A particular form. A headline. Copy for a Google Adwords ad. Something you can measure your current performance on, and measure a change to afterwards. If you're not high-volume, this might be harder to find than you think, but go through the exercise of figuring out what you can collect enough data on now, and what can wait to change until you can measure it later (or what you don't care about measuring). Then, you can take that one thing to a designer and say, "I'm looking to experiment with different changes to <X>. What are some things you think might be interesting to try? Anything goes." You're not asking them to do any work. You're not asking them to stake their reputation on the line. Your language ("experiment," "different changes (plural)," "interesting," "try," "anything goes") all says you're willing to respect their time and level of engagement and also that you're prepared to handle their advice properly, that you're not looking for a silver bullet, you're looking for all of the lead bullets. Say you ask for help with a landing page. A designer might suggest some copy or visual design tweaks for clarity (specific changes), but also suggest you test different headlines (without providing any), and also changing your format from short form to long form or vice versa (definitely without providing any content to you, but maybe they would be interested enough to sketch out some varied layouts). These are all great ideas. You should read up on A/B testing and test all of them in isolation, and also all together. You should test big dramatic changes even more frequently than you test small copy changes, because you don't want to be stuck looking for a local maximum. A UX person might also suggest ways to find out other things to test, research questions to ask, and suggest books or other resources so you can collect that data well. A startup that came to office hours that well-prepared would basically almost not need design help at all.

sandGorgon 12y ago

I see. Thanks for taking the time to write this out - and the answer to your question is yes: there are many places for people to look and review your code (not limited to http://codereview.stackexchange.com/ http://codereview.stackexchange.com/ and IRC). From all your answer(s), I get the feeling that UI/UX designers tend not to spend time for free because of one of three reasons: 1. a quick design help cannot be done and the alternative is too much effort to be justified 2. if you have an MVP before you are taking design help - its already too late 3. You really dont want design help - you want platitudes Let's say, for argument's sake, that I dont fall in any of these categories - built a MVP, took it to market, started generating revenue already. But I think I understand design (but am unskilled at it) and I can take a complete trashing of my current work. However, for certain startup-survival reasons, I need to make something incrementally better (both UI and UX). It's not the promised land, but just one teeny step towards it. How can I frame my request for help to not piss you off, not waste too much of your time and still not kill my startup ?

vitovito 12y ago
Your second-to-last paragraph is a good framing, and your final question is a good question. It's also a good example of why quick design help is hard, because there is no way for anyone but you to know what would be incrementally better or not for your users. All an outside designer can do is suggest things that might be interesting to test. You still have to try them and measure the change over time. A given designer has no specific knowledge about your market, and something can be well-designed in general but still have a lot of potential for market optimization. ( In fact, a colleague runs a service that is just that: he'll try something new for you, every month. https://draft.nu/revise/ https://draft.nu/revise/ ) Since only you have the data on your MVP, start there. Put together your customer lists. Make sure you understand their demographics and their usage patterns and why they purchased and what they were considering instead. Make sure you have ways to contact them for research interviews and if it's possible to go on-site to see them use your product live. Document how your competitors present and discuss their products. Then, pick something. One small thing. A landing page. A particular form. A headline. Copy for a Google Adwords ad. Something you can measure your current performance on, and measure a change to afterwards. If you're not high-volume, this might be harder to find than you think, but go through the exercise of figuring out what you can collect enough data on now, and what can wait to change until you can measure it later (or what you don't care about measuring). Then, you can take that one thing to a designer and say, "I'm looking to experiment with different changes to <X>. What are some things you think might be interesting to try? Anything goes." You're not asking them to do any work. You're not asking them to stake their reputation on the line. Your language ("experiment," "different changes (plural)," "interesting," "try," "anything goes") all says you're willing to respect their time and level of engagement and also that you're prepared to handle their advice properly, that you're not looking for a silver bullet, you're looking for all of the lead bullets. Say you ask for help with a landing page. A designer might suggest some copy or visual design tweaks for clarity (specific changes), but also suggest you test different headlines (without providing any), and also changing your format from short form to long form or vice versa (definitely without providing any content to you, but maybe they would be interested enough to sketch out some varied layouts). These are all great ideas. You should read up on A/B testing and test all of them in isolation, and also all together. You should test big dramatic changes even more frequently than you test small copy changes, because you don't want to be stuck looking for a local maximum. A UX person might also suggest ways to find out other things to test, research questions to ask, and suggest books or other resources so you can collect that data well. A startup that came to office hours that well-prepared would basically almost not need design help at all.