How to stop using images that suck

image of people

Does this photo suck? Let me know what you think in the comments section.

How do you know when you’ve found a good image for Web marketing? Many of us probably use the same criteria that Supreme Court Justice Potter Stewart used when trying to define obscenity: “I know it when I see it.”

An amusing but not exactly helpful benchmark.

I produce a handful of social media ads for one client, and images and image selection always seems difficult … because it is.

That’s why I was interested in a webinar recently hosted by on “Hero Shot” images that improve conversion. Presented by Angie Schottmuller of Three Deep Marketing, the webinar definitely helped me to identify and measure hero shots better.

I won’t give you chapter and verse on the webinar, but I will give you key takeaways on what I learned.

What makes a hero?

First, what do we mean by “Hero Shot” images? Well … it’s the dominant image used to demonstrate your product. It can be an image you use for a landing page, an offer page, in online ads or social media advertising, etc.

Schottmuller defined a hero image as:

“A credible photo or video of a solution that includes relevance, context and value and emotion to support, educate or persuade a customer.”

She then went on to define persuasion factors – but I’ll stick to the main takeaways that struck me as particularly helpful.

The first was a great analogy. Think of the words and keywords you use to attract attention to a hero image as a “scent trail.”

The keywords or text you use to draw interest are to initiate interest; you maintain interest on the page you direct visitors to – and you enhance interest via the hero shot.

[* Cool website for word association/ideation:]

This idea dovetails nicely with something other marketers have mentioned: the need for a relatively seamless continuity of images and styles when putting out a breadcrumb trail for your site visitor to follow.

The eyes have it

Another great imperative of hero shot images: pay attention to the eye path of the subject of the photo.

People naturally find pictures of people rather than objects more interesting. But trying to find or make the right images with people is very hard – specifically because of where the photo subjects are looking (or not looking).

People naturally follow the photo subject’s eye path. They will look where the photo subject is looking. So, is that to your offer, or a headline? Or is it off the page – to a blank space?

And another thing, according to Schottmuller: People making direct eye contact tend to disrupt the flow of what’s taking place on the page. That’s not necessarily a good thing or bad thin; the point is to use that tool properly. For instance, in a display or social media ad, you may want to use the full eye contact image.

On a page with a longer flow to it, full eye contact images may not be such a good idea.

Schottmuller also mentioned (of course!) authenticity – the age-old stock photo issue. But here I’ll add my own caveat, which kind of went unsaid by Schottmuller: a stock photo is perfectly fine, so long as it is relevant, conveys emotion and helps persuade the customer.

Emotional appeal

She also mentioned the importance of portraying desired emotions in hero shots. Don’t think of this as the correct emotion much as the emotion that leads to the logic and value of the offer.

This I found particularly helpful: don’t shy away from images that convey strong emotions. Key in on primary emotion such as: fear, pain, hunger, curiosity, love.

We may not want to admit it … but the right emotion can trump logic said Schottmuller. I think to people outside of marketing that may sound a bit cynical but … if you’re a marketer, you’ve probably seen this axiom play out in real life, in some way.

Remember what Flint McGlauglin of Marketing Experiments is fond of saying? “Clarity trumps persuasion” – and if you are able to portray that clarity with an image, so much the better, IMHO.

Another key point: Your customer is the hero … not you! This may seem self evident, but it has to be said. Because I really, really, really cannot stand to see generic images of “our employees” (you know the kind) on websites or in marketing.

These kinds of images just scream “It’s all about US!” or “We’ve all been taken over by pod people – for the love of God send help!”

Measure – and test!

And finally, what I found really helpful was Schottmuller’s recommendation to use a scoring system to help you decide on good photos. She recommended using a scale of negative one (-1) to plus 2 to rate images on 7 different criteria.

The scoring works like this:

  • -1: Negative potential
  • 0: Not quite conveys the point
  • 1: Somewhat
  • 2: Great

+1: for a successful direction cue

The criteria you can use to judge are:

  • Keyword relevance
  • Purpose clarity
  • Design support
  • Authenticity
  • Added Value
  • Desired emotion
  • Customer hero

This is a great system to use when you are weighing your image decisions. And although you may not necessarily find this makes every decision easier, it certainly helps you key in on what’s important.

And finally: test and measure your results. If you are using analytics of any sort, be sure to measure bounce rates on offer pages or pages using images. Of course, it’s all about conversion in the end!

Finally – the Bizarro lesson

Because I often think in terms of flip sides, here’s what I learned about hero images that don’t work. An image that is not helpful:

* Doesn’t have realistic emotions

* Doesn’t thematically or emotional match keywords in headlines, caption or teaser text.

* Doesn’t explain anything about the product’s value.

* Demonstrates zero concern/interest in the customers point of view,

Great job to Angie Schottmuller and

A Web marketer’s dilemma: the head and the heart

image of decisionData, data everywhere — and not a thought to think.

Is that what it feels like sometimes when grappling with analytics? Does it really inform the creative side of the equation? Of course, you know it does, but there always seems to be tussle between a “hunch” and a good solid data-derived decision, isn’t there?

If you have anything to do with Web marketing, you know that it’s a world increasingly awash in the promise of data-derived decisions. And yet … experience tells me that sometimes just plain old hard word works too. And so does a brilliant off-the-cuff hunch.

It’s a classic fight of whether you are following the head or the heart when trying to drive a result. Of course you are using data … but we all know at some point creativity and the ability to reach the human element matters, too.

So a recent post on the HBR Blog Network caught my attention with this headline: “What Data-Obssessed Marketers Don’t Understand” (requires free registration to view). It’s an interesting post that offers up an interesting thinking tool called “The Intelligent Brand Framework.” The two Gartner Inc. coauthors argue that the obsession with “big data” can obscure something basic: That data-derived decision making may be disconnected from what really motivates buyers.

Fair enough (though I suspect they are conflating “big data” with “analytics” … but that’s another story). They include a nifty chart, however, that I’m beginning to toy with for my own work. I must say it is a nifty tool — and it has a great name that makes me feel smart just for using it, too.

Or manipulated — I’ll leave that to your judgement.

So: I heartily recommend the piece and the tool. And check out the discussions at the end too. Good comments on this post from David Bloch, in particular.

[Photo Credit: Victor1558 via Compfight cc]

Sheena is a click farmer

image of The RamonesIn case you missed it, there’s been plenty of great articles and posts recently asking a very simple question about Facebook: How many Facebook accounts are bullshit due to fraud from click farms? A well-made video kicked off the recent spate of stories on the topic. Here’s a run down of what’s been in the news:


Veritasium on YouTube: Facebook Fraud 

Washington Post: This blogger paid Facebook to promote his page. He got 80,000 bogus Likes instead

Salon: Facebook’s black market problem revealed

USA Today: Want fans? Hire a social media ‘click farm’

This of course leads to the natural follow-up question that any marketer should ask: So how much bullshit am I paying for when I buy an ad on Facebook?

I really loved the “Facebook Fraud” video (the first link above, by real-life science geek Derek Muller). However, there are some very big caveats to his video which need to be mentioned. Muller starts off mentioning an experiment run by a BBC reporter back in 2012. The reporter set out to determine what a “Like” was worth by building a Facebook account for something called “Virtual Bagel.” Then, he used Facebook’s ad platform to buy $100 worth of “Likes” … in countries where known click farms come from (gee – wonder if that affected results?).

The lesson here is fairly obvious. If you don’t target your ads (which is fairly easy to do on Facebook ad manager settings) I would imagine that of course you are attracting a lot of bottom-feeders. That’s lesson number one (and keep track because there will be a quiz). Facebook in 2012 since then said it deleted millions of “fake” accounts from click farms — something Muller follows up on by reproducing (more or less) the “Virtual Bagel” experiment.

What he discovered, however, is that nearly all of those new likes he got from a recent experiment were not engaging with his page. So … nothing seems to be solved, and may indeed be getting worse (according to Muller). In fact, it gets more nefarious. Many of these new accounts didn’t appear to come from known “click farm” countries. They seemed to be coming from the US. But there was something odd about these Likes … these accounts “Liked” way too many things to seem natural — and the things these “people” liked were odd.

My own experiment

Recently, I conducted some testing of Facebook (and other) ads for a client. I’m not going to divulge the results, but I will say this. Of the recent Likes we gathered, I decided to look at a representative sample to see if I could find accounts that screamed “click farm” — clearly.

I did not find a torrent of fake accounts. Not by a long shot. You see, I took the trouble to carefully target the ads.

However, I certainly found between 10% and 20% that were suspicious. A few fairly screamed click farm — which seemed to represent about 12% of the total (I looked at a sample of my new Likes — not all of them). Either way you cut it, paying for obvious fraud is not acceptable.

One of the new Likes my client gathered was from a person I’ll only identify as “Sheena.” Because there is a small chance — very, small — that I’m wrong, I won’t identify the account completely. But let me tell you a bit about “Sheena”…

Sheena certainly likes a lot of things. And yet, the only thing on her profile is that this blonde-haired twenty-something is female. Good to know! Oh, and according to her timeline … she changed her cover photo. There are no other posts to see.

Oh, and she sure likes a lot of things. To be specific … 33,000 things. Just to put that number in perspective, if you “liked” five things a day, for every single day of the year, rain or shine, come hell or high water, it would take you 18 years to hit that number. Or let’s say 20 things a day, for every single day of the year for 4 and a half years … Possible, but not plausible.

And it’s not just that Sheena likes so many things (and she’s so young and … blonde … too!) This little OCD like-monkey has quite the unusual taste in what she “likes.” It seems, for instance, that she “Likes” 10,000 restaurants. Which is quite the feat. I’m sure I’ll get to know “Sheena” better when we all see her on a future episode of “My 600-pound life” on the The Learning Channel. And it’s not just that she likes so many restaurants … she even likes every location of a franchise! Isn’t that neat-0!

She also “Likes” the Chicago Blackhawks AND the Montreal Canadiens. If you know hockey, you know how ridiculous that last sentence is. That’s like saying you’re a vegetarian who just LOVES to pick out their own baby cow to kill to make veal patties.

The sad reality is, I don’t really know if Sheena is a click farmer or not. I do know her account fairly screams it. My concern is that … I can’t tell. And that should worry Facebook a lot that people putting money on the table can’t really tell what’s fake and what’s real.

(And yeah, I supposed I should have named the blog “Is Sheena a click farmer?” … but “Sheena is a click farmer” sounds too close to “Sheena is a Punk Rocker” by the Ramones. And I like the Ramones. Suck it Robin Thicke)

I will say this: if you don’t target and test, you are opening yourself up to manipulation by fraudsters. Maybe that’s the price of playing this game. Nonetheless, I doubt there are very many businesses that contain an asterisk in their billing statements that reads “*Your results may vary. Sometimes you’ll be paying for a lot of fraud.”


Photo credit:By Plismo (Own work) [CC-BY-SA-3.0], via Wikimedia Commons