10 May 2010
Human Photos Double your Conversion Rate
Editors Note: In his second article for Think Vitamin Paras Chopra looks at how human photos can increase dramatically increase your conversion rate. For further reads check out Paras’ first article “The Business Case for A/B Testing” and ways in which you can “Designing for Conversion Experiences” by Chrissie Brodigan.
It is a well studied scientific fact that in a scene or image full of different objects, one’s attention is subconsciously drawn to human faces. Even toddlers and monkeys look towards faces for a much longer duration than they’d look at other objects.
Independent Tests Yield Same Result
I have analyzed two A/B testing case studies of Visual Website Optimizer users and have concluded that the sure fire way of catching visitors attention is having human photos on a website.
What’s interesting is that even though both tests were independent, they (surprisingly) tested the same aspect and arrived at the same result. This gives me confidence that there is something special about having human photos on a website which increases conversion rate instantly. Let’s have a look at the two studies.
Case Study #1
Medalia Art sell Brazilian and Caribbean art online and they recently got pleasantly surprised when photos of artists on their homepage increased visitor engagement by 90%. This wasn’t their first A/B test as they keep playing with their homepage. In the past, they successfully reduced bounce rate by 20% by testing the placement of their promotional message.
As can be expected from an online art shop, they showcase paintings from three famous artists on their homepage. This time, for no good reason they decided to see what happens if paintings are replaced by the artists’ photos. I tried to probe Medalia Art for reasons why they came up with this idea but they said it was a completely random choice to test paintings v/s photos.
The goal for this A/B test was to increase visitor engagement (defined as a click on any link on the homepage; it is the inverse of the bounce rate). They didn’t use sales as a conversion rate as many of their sales happen after interaction on the telephone (and a lot of haggling) with the customer.
Here are the screenshots of the control (paintings) and the winning version (photos).
Control (Paintings)
Variation (Photos)
Test results are as follows:
| Variation | Conversion Rate | % Improvement |
| With paintings (control) | 8.8% | - |
| With photos (variation) | 17.2% | +95% |
Clearly, version with artists’ photos had a much better conversion rate that the version with artists’ paintings. In fact, the winning variation almost doubled the conversion rate on the homepage.
Case Study #2
EmptyMind is a web analytics and optimization blog by Jason Thompson. There is a large phone icon in the blog’s sidebar which links to his contact us page. Unlike Medalia Art who did the test just to check what happens, Jason had a reason for running a phone icon v/s his photograph test on the site.
While attending a summit, he got several comments from other people that he didn’t look like his (then) Twitter icon. So he thought it would be best to have a single photo representing him across all the social network sites he uses. Testing the photo on the blog was just an extension of that thought.
The goal for the test was visits to Jason’s contact us page and here are the variations and their performance:
| Control | Variation |
|
|
| Conversion Rate: 3.7% | Conversion Rate: 5.5% |
| Increase in Conversion Rate: +48% | |
Here is what Jason said about the test results:
People want to connect with other people emotionally, the photo makes that emotional connection so much easier and as the test is proving, drives people to the contact form more than a nondescript icon.
The Key Lesson
Both case studies independently arrived at the same (statistically significant) result that human photo(s) increased conversion rate. This proves that there is certainly something magical about human photos.
I am going to try this test on Visual Website Optimizer homepage (perhaps using customer photos for testimonials). If it is relevant for your website, I recommend you try an A/B test to see if a human photo can do wonders for your business too.
If you try out a test involving human photos and get positive results, then we may have discovered a great formula for increasing conversion rates!
Enjoy this article?
If you liked this article, feel free to re-tweet it to let others know. Thanks, we appreciate it :)
We love .net Magazine
We're big fans of .net so they've hooked us up with a rad deal: Save 45 percent on your subscription.
Andrea
# May 10, 2010 - 7:56 pm
Very intresting post, I will try to make a test with and withouth photo on contact page.
Thank you :)
Paras Chopra
# May 12, 2010 - 8:38 am
Thank you for liking the case studies :)
Armin C.
# May 10, 2010 - 9:11 pm
I knew that faces drew users attention, but I didn’t realise just how much. Thanks for this tip ;)
Paras Chopra
# May 12, 2010 - 8:39 am
Glad you liked it. But don’t consider it a tip and apply it as it is. Do A/B test it first. May or not work for all websites.
Steve
# May 10, 2010 - 10:59 pm
I do believe when it is your personal portfolio which you use for job seeking the photo you use to represent yourself needs to be geared toward employment.
Nothing too formal but definitely watch the facial hair and urban clothes… I have seen many people be turned down as candidates just based on their ‘trendy’ folio photo.
Paras Chopra
# May 12, 2010 - 8:41 am
Yup, going too far with this concept may also hurt. Of course, you wouldn’t put your facebook photos on your personal portfolio for employment :)
Mike Healy
# May 11, 2010 - 3:24 am
I wonder if stock/custom photography makes a difference. I have a feeling most people don’t connect to the super-cliched business stock photography.
Paras Chopra
# May 12, 2010 - 8:54 am
Stock photography v/s genuine photos. Perfect case for A/B testing ;)
Matt
# May 27, 2010 - 7:27 pm
And we all know what they are! e.g. businessmen handshakes, attractive women in headsets etc.
I’ve actually thought about using them myself in times of desperation, but I’ve managed to resist.
Hitesh Mehta
# May 11, 2010 - 6:59 am
Thanks for the detailed article. Very useful.
Robin Carlsson
# May 11, 2010 - 10:34 am
Very interesting. I can see a problem with web designers as myself, that we tend to use more design elements than photos cause it often makes a easier color match with the rest of the site.
Joel
# May 11, 2010 - 11:12 am
Interesting results – would be nice to know how many total visitors there were for each test to get an idea of the margin for error :)
@Steve – depends how corporate the jobs you’re applying for are… I like to think of my facial hair as self-selecting, it keeps me away from jobs where I’d end up slitting my wrists with post-it notes
Paras Chopra
# May 12, 2010 - 8:59 am
I can assure you that results were statistically significant. Revealing exact number of visitors isn’t possible because that is a data point a lot of our users choose not to publish.
Ben
# July 5, 2010 - 9:49 pm
Give us the confidence intervals then. Don’t act like it’s a precise number – you can never statistically find the true mean.
Steven Wilson
# May 11, 2010 - 3:23 pm
Interesting article, worries me though as I’m the least photogenic person I know! never the less I’m going to give it a go with my new website design.
Katherine
# June 25, 2010 - 6:38 am
Use an image of yourself as a child or a baby (assuming a light touch suits the situation). I bet you were very photogenic then.
TheAL
# May 11, 2010 - 6:26 pm
Very interesting read. I won’t be cramming photos of cute babies into my navbar just yet, but figuring out how to tastefully use photos to get users to do certain things sounds like a fun challenge, especially for people who like to minimize the use of raster images on a site.
Paras Chopra
# May 12, 2010 - 9:02 am
Well, what you describe is a debatable issue. A lot of times A/B test results produce an ugly version as a winner. As a designer would you use aesthetics over better results?
TheAL
# May 12, 2010 - 7:08 pm
And therein lies the debatable issue. Ask me five years ago. I would have said, “Heck yeah!” Being educated as a systems analyst slash programmer, graphic design wasn’t always my forte. UI in general, maybe, and even then only to an extent. I’ve really just been self-teaching design over the last few years. I definitely care about a site’s conversion rate, but using photos in a design can still be tricky.
What should we strive to achieve? Ugly w/ a high conversion rate? A designer’s wet dream with a poor conversion rate? An acceptable compromise between? I’d like to hear some varied opinions.
John Harvey
# May 11, 2010 - 6:38 pm
Great article, I have just been testing my latest site and have used human and animal images in each of the sell pages, lets see if your theory works. Thanks again. Regards John.
Paras Chopra
# May 12, 2010 - 9:04 am
Fantastic! I will be curious to hear your results. Maybe you would want to throw in a version with generic icons just to see if the theory really works?
Steinar Knutsen
# May 11, 2010 - 7:10 pm
Thanks for the results from this study. I am definitely going to test this out. I use pics in nearly every blog post and this might by a useful tactic to draw in readers.
josh
# May 11, 2010 - 7:47 pm
Very interesting indeed. Especially the contact button vs. a pic and the word contact.
Lucas
# May 11, 2010 - 7:53 pm
I wonder if this applies to using generic stock images with people instead of using graphics or illustrations for your site. Some stock images are powerful but there’s a lot of girl-with-headset for contact page going on out there. Would there be any difference if there’s a phone icon or something?
I guess it needs to be tested since in the cases above the images represent the actual artist or the actual contact person.
Paras Chopra
# May 12, 2010 - 9:06 am
Maybe there is a reason why stock photography is used so often. We all are subconsciously attracted to pretty human faces, so depending on the website, it never hurts to test stock photography on a site.
Ali Ale
# May 12, 2010 - 3:56 am
Actually I never use my picture to all my bio profile or blog. After read this post, may be I’ll take my picture out.. Thanks!
Gerben van Dijk
# May 12, 2010 - 8:05 am
Very interesting read-up, thanks again!
mintocity
# May 12, 2010 - 10:55 am
Awesome article, really inspiring… thanks for sharing
ab3design
# May 12, 2010 - 9:01 pm
Interesting, but it strikes me that the example using the artists might be influenced by other factors than some innate response to human photos – namely the fact that when the image used to represent the link was an example of the artist’s work, this allowed users to make an early judgment as to the value of clicking that link – (ie, “hmm… don’t like that painting, don’t need to see more, so I won’t click here”) whereas the image of the artist gives no information as to what that person’s art looks like, and following the link is required to get that information. I’d like to see the results of this test with a control – say, with no image associated with the link, or a random image image that is neither a photo of the artist or an example of the work.
Paras Chopra
# May 13, 2010 - 7:17 am
Ah, an interesting observation. Quite valid one too. It may certainly be possible that visitors were making judgement on paintings. Shall ask Medalia Art if they can follow up the test with your suggestion.
Ollie
# May 16, 2010 - 7:23 pm
Agree with Ab3 on this. Art is so subjective, it’s more likely that someone won’t like a painting or style than they will like it. By substituting the artist’s photos, the infromation people want is just one level lower in the site heirarchy, forcing a click. Bounce rate is the wrong metric to watch, perhaps?
Also, not sure that case study 2 is all due to the photo. The control design clearly showed a phone icon, where the amended design had just Contact Jason with his photo, without specifying or hinting at the contact method. Is it possible that people thought in the control situation that they could only contact by phone, which many people might not have wanted to do? But in the second situation, people may have expected that they could contact by several methods, so clicked more often.
Paras Chopra
# May 17, 2010 - 4:00 pm
Hi Ollie,
Yes, bounce rate isn’t the best metric here. But for Medalia Art a sale is a very complex process involving lots of phone calls with the buyer, so tracking sales for them in the A/B test is a bit difficult.
Regarding your comment on case study #2, I don’t think the phone icon may have any effect. In fact, it is an overused icon so people may have developed a blind eye for it (just like many people are blind to ads). A photo presents to people something they haven’t seen before plus it provides an emotional connect with the blogger! That way, readers also realize that there is a human behind all these posts :)
-Paras
Kaffebønner
# May 18, 2010 - 8:14 pm
Nice article – thanks!
philip foeckler
# May 20, 2010 - 5:00 pm
what about puppies?
Matt
# May 27, 2010 - 7:25 pm
Very interesting! I’ll definitely be incorporating this technique into my corporate website when I give it a big overhaul in the next couple of months.
I think human photos must give a more trustworthy impression of the business / website owner, rather than it being a bunch of people who want to hide their identities behind their computer screens.
Jason Sew Hoy
# June 1, 2010 - 3:48 am
I’d love to see some case studies around the use of human faces to increase conversion when the goal is to drive sales of a product or service.
I’d suggest this isn’t as predictable as measuring clickthroughs to an “about us” page and it’s debatable as to whether human faces would improve conversions or distract people from the conversion funnel.
Guess there’s only one way to find out!
Steven
# June 24, 2010 - 2:57 pm
Great post. And very timely for me as I have a client that does not want his own photo posted on his site due to the industry he is in. They have suggested using local images of the area and generic stock photography.
I have pushed the “personalization” and wanting to use their own staff.
Now I can show them your article and at least try to get some quality stock photography with people.
John Hyde
# June 30, 2010 - 11:56 am
It’s always good to see results from tests.
Interesting that all the photos are looking back at the viewer – it would be interesting to see tests where the person in the photo was looking at something else – eg towards a product or a form.
Reklama
# July 9, 2010 - 7:15 am
That’s really impressive, I’m gonna try it. It’s really good to see articles with results of concrete actions that increase the conversion rate.
Thanks and keep it that way.