Austin McCraw

This Just Tested: Stock images or real people?

April 8th, 2011

In our most recent Web clinic on optimizing leads, we quickly reviewed a recent case study in which two banner images were tested – a generic stock image vs. an image of a real person. This experiment led to more insights than we had time to cover last week; so, I thought I’d give it a little more room to breathe here on the blog.

CONTROL: Who doesn’t love a generic smiling lady?

If you haven’t yet watched the Web clinic replay, the company (blurred intentionally) we were working with in this experiment was a consumer credit counseling service offering free debt consultation. Their homepage had been the focus of many previous radical redesign tests, but for the scope of this research project, we were focusing on one particular issue: The main banner image.

Currently, the homepage utilized a generic stock image of a random lady wearing a headset, basically attempting to bring some reality to a free debt consultation. The main draw (as I see it) is that a really friendly attractive lady is standing by to help you with your credit challenges – why not give her a call today?

But who is this lady? How can she help me? And why in the world does she keep smiling at me?

Your customers are pretty savvy. They know Smiley McHeadset doesn’t really work for your company. They know she’s a paid model and is likely smiling in an ad for a bank and a billboard for a credit card as well. Do you think that deepens the trust with your company?

Or as David Meerman Scott said at this year’s Email Summit, “The problem with the B2B happy multi-cultural conference room with computer shot is that it has become a cliché. It is world-class, cutting-edge, mission-critical visual gobbledygook. Just like written gobbledygook, this kind of image is so overused to have become meaningless.”

TREATMENT: Putting a face with a name

The question arose from our team, “Would visitors respond more to a familiar face over this generic one?” The thought was that the higher relevance created by a recognizable image would more closely tie into the motivations of the ideal prospect.

So, to test this hypothesis, we pitted the happy-go-lucky anonymous phone rep woman against the popular, reputation-wielding company founder. I guess you could say it was a “head-to-head” competition (ok, sorry I couldn’t resist). Now this new image represented a real person, with real credentials, and was highly recognizable for channels coming to the home page (television ads, news articles, etc.).

Side note: Notice anything else different about the images. Yep, that’s right, gender. Might gender play a role here as well? Possibly. (So might weight, hair color, attractiveness, etc.) But just so you aren’t bogged down by this question, this lady had been tested against other male generic stock images, and was the current reigning champion (no offense to all our male model blog readers out there). This test treatment, however, was the first in which we tested a stock model versus an actual person.

RESULTS: Familiarity Breeds… Conversion

So what were the results? Well, Mrs. Generic finally met her match. It appears that an attractive smile is not a match for a good name. Overall, the familiarity hypothesis held some water. When the recognizable image of the founder was used, visitors were 35 % more likely to sign up for a free consultation. Remember, this is a 35% lift on top of many other previous gains in the testing-optimization cycle.






INTERPRETATION: Quick! Sell your iStock credits before this post goes viral

So what are the implications of this test? Are all of our generic stock images worthless? Should you trade all your iStock credits in for a new camera? Well, not exactly.

Stock images are not the enemy here – irrelevance is. You could hypothetically take your own irrelevant picture and it wouldn’t make a hill-of-beans difference – except that it probably wouldn’t look as good. The main point here is that how we choose images matters. This is true for all graphics, and the following principle is key.

KEY PRINCIPLE: An image is only as valuable as the value it communicates

This is how we must think when choosing images. We shouldn’t just pick images we think make our site look prettier, but rather, we should choose images because they say something about the value of our offer. So, if you use a stock image of a smiling lady, make sure it is communicating value. If you use a picture of a middle-aged balding CPA, make sure it is communicating value. It really doesn’t matter where your images come from, so long as they communicate relevant value.

To help bring the rubber to the road, the following questions can serve as a grid for finding the most effective image. You can see how the images above fit within this grid and why potentially the image of the founder generated the most response.

Image Assessment Grid (Click to expand)

(Note: we also discussed these questions more deeply in the article entitle Images vs. Copy.)

AUDIENCE FEEDBACK: What image would you test next?

Before concluding, I wanted to ask you if there’s an image that you think might perform better than the picture of the founder. We have discovered a picture that is more RELEVANT, but is there an image that might bring more REALITY (per the grid above)?

Maybe you can make a convincing case for what kind of image our research team should test next…


Additional Resources

Images vs. Copy – How getting the right balance increased conversion by 29%

Converting Leads to Sales – How one company generated $4.9 million in additional sales pipeline growth in only 8 months

Surprising Wins from 2009: Using insights from an uncertain economy to drive 302% growth

Online Ads Tested — How Matching Ad Design to Context Improved Conversion by 127%

Austin McCraw

About Austin McCraw

Austin McCraw, Senior Director, Editorial Content, MECLABS Institute From Web clinics to certification courses, Austin works behind the scenes to bring our research analysts’ discoveries to our audience in clear and creative ways. Before joining MarketingExperiments, Austin was the Promotions Director for a PBS affiliate in Gainesville where he was in charge of publicizing local and national programming. A graduate of the University of Florida’s College of Journalism and Communications, Austin specializes in new media and video production by day and producing family memories with his wife and two children by night.

Categories: General Tags: , , , , , , , , , , ,

  1. April 8th, 2011 at 04:17 | #1

    In Germany many client briefings contain the sentence:
    Use only license free stock pictures.
    Thanks to your article I can tell them, why this might be the wrong way.

  2. April 8th, 2011 at 05:41 | #2

    Perhaps you could test a picture of a happy family/man/woman with a headline like “They/She/He got relief from debt. And you?”

  3. April 8th, 2011 at 10:52 | #3

    I would test a photo of an actual customer along with a testimonial.

  4. April 8th, 2011 at 15:51 | #4

    I guess I could have predicted this, but you never know until you try it.
    Picture your website as a meeting with the client. An analogy: it’s like an in-person meeting. They are trying to get to know you better and you put up a cardboard cutout of a beautiful person in front of you. They hear your voice in the conversation, but it’s awkward because they can’t see the real you. It’s very impersonal.

  5. April 8th, 2011 at 23:19 | #5

    I think the issue is a matter of relevance combined with creative and unique imagery. The difficulty with the current smiling stock photo woman is that the actual customer service representative would still (hopefully) be smiling like that when they are providing the actual service (discussing debt). Talking on the phone is a rather generic action to display in marketing and advertising. Compare this with an architecture, graphic design, or law firm, where the services incorporate more specific visual elements than just talking on the phone. It’s easier to display the relevance of these latter examples in a more unique way.

    One issue may be that the headset lady is simply a cliche, knee jerk way of articulating the service. One still might be able to generate results on these websites with a stock photo, but it may require some creative thinking as to how to visually communicate the debt consultation, without limiting the imagery to what the consultant actually physically does when providing the service.

    Additionally, you might want to also consider testing a photo of founder interacting/meeting with a customer (in person). This would realign the response away from “Why is this person smiling at me” to “Okay, now I understand a bit more about what service is being offered by whom.” I wonder how study results would differ across different industries and services.

  6. April 10th, 2011 at 15:12 | #6

    What a fantastic post!!! You really did the process well here, and it paid off big time. The company I used to work for is widely advertised on the radio and could greatly benefit by incorporating the CEO’s pic into their online advertising, since it’s his voice on the radio. Great stuff!

  7. April 13th, 2011 at 10:47 | #7

    I´m so glad you cited the debt relief sector as an example of this herd practice.

    ClearDebt a UK company tried a similar experiment and found ´middle aged realistic slightly haggard and stressed” lady worked (converted) better than “attractive model” image. This was also we believe in part due to the direction of the gaze (ie better when directed to the call to action instead of out to the site visitor). The performance increase in our case was 27%.

    This could further be improved as you infer by using “real person” imagery for the click to chat imagery instead of over reliance on stock libraries.

  8. April 13th, 2011 at 12:47 | #8

    ClearDebt in the UK did a similar AB test and found the haggard middle age lady outperformed the younger attractive subject by 27%. It was also we believe because the direction of the gaze of the better performer was towards the CTR and not, as typical, towards the screen, ie towards the page viewer. Nonetheless authenticity, especially in the debt sector is increasingly important and the days of stock imagery, especially in the social age are hopefully over.

  9. Marla Hughes
    April 17th, 2011 at 07:43 | #9

    @Paul Gailey
    Excellent real world example of the article’s principles taken one step farther. As I’m a ‘haggard and stressed’ cleaning service owner, pictures of me cleaning should sell my services like hotcakes. 🙂

  10. April 18th, 2011 at 13:19 | #10

    Thanks for the thought-provoking data and post!

    It seems to me that there are really several factors.
    – Relevance, certainly – do I believe that this image is directly connected with the offer?
    – Believability – is the person I’m talking to really going to be smiling?
    – Credibility – do I believe this person has integrity and do I trust the message they are representing?
    – Resonance – does this image convey the feeling I expect and want to have when I work with this company?
    – Veracity (for lack of a better word) – is this image a window into the experience I’m going to have working with this company?

    Websites done well convey the sense of being a lens into the business, so that the visitor understands who they are dealing with and what kind of experience they can expect to have. I’m sure that’s part of why blog-centric sites are so popular as marketing tools now. Read enough blog posts and you eventually know how the company “thinks.”

    Since we’re about to overhaul our site, this is timely and useful. Thanks!

  11. May 6th, 2011 at 15:07 | #11

    What would have been a more interesting test would be to test generic woman against a photo of one of the real customer service reps. Not everyone has a recognizable face that can be added to their site. Relevance is important of course, so should you use a stock “relevant” image or a “real” relevant image? Your test did not answer this question.

  12. May 9th, 2011 at 11:45 | #12

    Why no facebook share button? Are we a part of an experiment? Good article – I’d like to see what happens when you “name” the model

  13. May 9th, 2011 at 16:34 | #13

    This girl works for every company — — from banks to doctor’s offices to industrial supply companies.

  14. June 2nd, 2011 at 12:02 | #14

    You tested two variables here: ubiquitous stock vs. unique picture, and anonymous vs. recognizable person. If you had chosen a random male employee from the company, who knows whether you would have gotten the same result.

    But as you pointed out in the post, the key, eventually, is relevance.

  15. August 14th, 2011 at 22:35 | #15

    This is very helpful data – A lot of clients have a library of stock images that they use with their marketing collateral and it’s often a battle to get them to realise how cliche it looks – thanks for the help in putting this message across!

  16. November 24th, 2011 at 23:57 | #16

    Agreed with Aaron. I’ve had this argument with several people now. The worst example are dentists…I’ve seen the same generic brunette with a white toothy smile as part of the sign in the front window of two dental offices and as part of the front page of four websites. How’s that supposed to distinguish any of these dentists?

  17. February 17th, 2012 at 08:41 | #17

    Totally agree. That’s the reasons why I choose to put my (big) face in the homepage of my website 😉

  18. April 25th, 2012 at 18:10 | #18

    Awesome post on Stock & Real Photographs for the site. Instead we can use semi realistic rather fake pics in which people will be more interested. And it rather depends on the Profession or sector.. Thanks.

  19. August 3rd, 2012 at 19:43 | #19

    As a stock photographer I find this post very interesting. My goal is to shoot ‘real people’ and even though there will always be a need and use for real stock photos, they can’t ever replace the actual face of a company. Figuring out which one should be used in which situation is the trick. Obviously A / B testing is needed in most cases – but very revealing that a middle aged male owner can out perform a pretty smile. People do clearly like and trust familiarity

  20. March 20th, 2013 at 20:23 | #20

    After exploring a few of the articles on
    your blog, I really like your technique of writing
    a blog. I bookmarked it to my bookmark site list and will be checking back soon.
    Please visit my website as well and let me know what you think.

  21. March 17th, 2014 at 20:25 | #21

    As a stock photographer I find this post very interesting. My goal is to shoot ‘real people’ and even though there will always be a need and use for real stock photos, they can’t ever replace the actual face of a company

  22. March 17th, 2014 at 20:27 | #22

    The company I used to work for is widely advertised on the radio and could greatly benefit by incorporating the CEO’s pic into their online advertising

  23. July 14th, 2014 at 10:45 | #23

    @Joshua Camp

    I agree, put a name below the model or use one of the long retention reps pictures with her name and commitment to helping the customer with financial problems and I think it will raise the bar on your results. 🙂

Comment pages
1 2 6542
We no longer accept comments on the MarketingExperiments blog, but we'd love to hear what you've learned about customer-first marketing. Send us a Letter to the Editor to share your story.