Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Does Your Product Logo Actually Matter?

kalzumeus.com

Post reply on HN
Print view View original

31–40 of 54 posts

Re: Does Your Product Logo Actually Matter?

#31

Post by tptacek »

patio11 wrote: post #29
tptacek wrote: post #22

Earlier quoted context omitted.

Try. But before you do: are you sure you know what Patrick's business actually is ? It looks like "Hello world" hooked up to a random number generator... but it isn't. Here's a tip: when you've figured out what it is Patrick is actually doing, you'll see why it's kind of silly to consider competing directly with him.

Thomas is presumably artfully phrasing "If you can do organic SEO and AdWords well, you can make a lot of money, so why the heck would you enter bingo cards" to avoid hurting my feelings. He isn't wrong about the money, though my feelings are pretty much impervious. I mean, I sell bingo cards to elementary schoolteachers: it is highly unlikely you're going to say something I haven't heard in the last four years. (Tho…

I think you're stripping what you do down too far.

Organic SEO and Adwords are tactics, but your business is a strategy that addresses the needs of a very narrow group of people who will pay to make study exercises (and occasionally family entertainment) easier and more effective. There are lots of other tactical moves you make, like figuring out how to demo your product, moving customers to the web version of the app, keeping metrics on task completion inside the app, etc.

But, yeah, the underlying theme here is that if you can execute these tactics and knit them into a coherent strategy, there are thousands of other niches you can address besides "people who will pay money so they can easily spend class periods playing bingo". Why waste time fighting over the bingo pie?

I'm not even a little bit worried about hurting your feelings. ;)

Re: Does Your Product Logo Actually Matter?

#32

Post by tptacek »

DotSauce wrote: post #8

Commented on blog with some personal logo design tips. I think the bigger problem here is that the rest of his site looks straight out of 1998. The half-decent logo may be the only thing lending it any credibility. :P

If you too would like a reason to donate some money to Teach For America, you could suggest a series of design changes and bet that they'd A/B test significantly better with his audience than his current design. I'm guessing that Patrick wins either way on bets like this, and so do we (since the writeups are great). So... what's the next thing you'd change?

Re: Does Your Product Logo Actually Matter?

#33

Post by patio11 »

nollidge wrote: post #28
patio11 wrote: post #23

Earlier quoted context omitted.

OK, quick stats 101: The hypothesis isn't that A > B, it is that A = B (the null hypothesis). 99% certainty is to reject the null hypothesis, leading to the conclusion that A is better. The second experiment you've outlined is gibberish, which involves a series of words spelled similar to ones you can find in a stats book, connected in a manner which leads to nowhere.

What's gibberish about it? I'm saying if you want to know if off-the-rack logos are, on average, better than custom-designed ones, you have to compare lots of them, not just one example of each. What's the problem with that?

Let's not talk A/B tests and talk medicine for a moment, shall we? Here we have aspirin, here we have a placebo. There isn't a population of aspirin pills and a population of placebo pills. There aren't many different populations of placebo pills. The populations are of people. We're trying to make inferences about the population of people exposed to aspirin versus the population of people exposed to a sugar pill, but since it is not possible to bring all of them in, we make those inferences based on random samples of those two populations.

Now here is where I start getting murky: results for the particular things we're interested in are binary: the aspirin worked or it didn't. You converted or you didn't. The Central Limit Theorem says that, if you sample n individual unconnected random binary events from a population, and you keep repeating that sample, the distribution of your measured percentage of success will approximate the normal distribution, centered around (with the mean on) the true population statistic.

That is important for a couple of reasons, but the most immediate one is that it gives us a statistical figleaf to hide under for using two samples to make inferences about whether the population mean for the first distribution is highly likely to not be the same as the population mean for the second distribution. i.e. whether the A and the B are likely to be materially different.

So let's walk this back to where we were talking about potentially testing multiple different types of sugar pill. Do you get why that is sort of incoherent in this framework? There are key assumptions like "independence" that two different types sugar pills surely do not share, but I don't know if you can even get to that objection before you deal with the notion that there is no coherently cognizable population made out of sugar pills.

It is 4 AM in the morning. I am not sure that is stats class talking and not sleep deprivation. This is probably covered in a good Stats 101 book -- if you have further questions, they're likely more coherent than I am. (Or ask btilly -- he has a gift for explaining this.)

Re: Does Your Product Logo Actually Matter?

#34

Post by jnorthrop »

Honestly I'm really surprised that the logo made that much of a difference in sign-ups. To me this indicates that you should try even more logos to optimize your sales. Maybe both logos are performing poorly and a "good" one might double your customers.

If there is anything to take from this experience it isn't that bespoke is better than generic designs -- its that the logo matters, potentially a lot.

Re: Does Your Product Logo Actually Matter?

#35

Post by chaosmachine »

It seems blindingly obvious to me that getting rid of the bingo card would hurt sales. It doesn't matter what you replace it with, nothing is going to communicate "I have bingo cards" better than a picture of a bingo card.

Here's a much more interesting question: Have you found the optimal bingo card? What if you ditched the blue circle, and just made the card bigger? What about the font/colors? Can you get a 1% increase in sales just by iterating your existing logo?

Re: Does Your Product Logo Actually Matter?

#36

Post by ced »

Patrick, it sounds like you're performing "orthodox" statistics. That's likely to be good enough for this problem, but have you considered Bayesian statistics at all? Your conclusions could be very different.

Re: Does Your Product Logo Actually Matter?

#37

Post by gojomo »

What if some of the visitors had visited before -- or seen the prior logo before, on Patrick's blog or another story -- but from another browser or where the cookies had expired/cleared? Then, seeing the second logo on a followup visit would have a 'slightly off' look that could deter signup. ("Is this a knockoff site?")

Re: Does Your Product Logo Actually Matter?

#38

Post by angumagu »

I'm concerned on two fronts: 1) if quality of visual design is the thing being tested, then it might be more useful if the objective information in the two logos were the same. As it is, the difference here could be entirely attributed to the bingo card image being in one logo and not the other. (I know, the choice to put it in is a design decision, but it's useful to distinguish between testing the design/look-feel and the information in the logo) 2) I don't see in the write-up where brand confusion is accounted for here. What percentage of those shown the new logo had seen the previous one on a previous visit, and turned away because they thought they had the wrong site, or had otherwise diminished their view of the brand because of confusion?

Re: Does Your Product Logo Actually Matter?

#39

Post by nollidge »

patio11 wrote: post #33
nollidge wrote: post #28

Earlier quoted context omitted.

What's gibberish about it? I'm saying if you want to know if off-the-rack logos are, on average, better than custom-designed ones, you have to compare lots of them, not just one example of each. What's the problem with that?

Let's not talk A/B tests and talk medicine for a moment, shall we? Here we have aspirin, here we have a placebo. There isn't a population of aspirin pills and a population of placebo pills. There aren't many different populations of placebo pills. The populations are of people . We're trying to make inferences about the population of people exposed to aspirin versus the population of people exposed to a sugar pill, b…

I've got Stats 101 knowledge, biostats in fact, but perhaps I've misapplied it here. I'm at work now, so I'll maybe revisit my thinking later.

Re: Does Your Product Logo Actually Matter?

#40

Post by potatolicious »

tptacek wrote: post #26

I think I actually "chose" this logo (in that it was one of 5 logos I suggested to Patrick when we started this bet). My hypothesis was that any professionally-designed logo, generic or otherwise, would outdo Patrick's bespoke logo, which positively screams "April 19th, 2002!". ... and I was wrong. Teach For America an acceptable charity, Patrick? Someone in this thread suggested that this shows-t'-go-ya that geeks s…

I think you're underestimating the value of authenticity, especially within the right demographic.

Bingo Card Creator (correct me if I'm wrong) is primarily targeted at a demographic that is friendly to small business. An attempt in this market to appear like big-faceless-megacorp will, as this study proved, hamper your results.

The old logo may not look terribly good, but what it says to me is: authentic small business run by some dude who sure appreciates my business.

It always surprises me how many small companies try so hard to appear big, when in fact there's so much value in projecting that of the underdog.

Post reply on HN