Creative Analytics for More Conversions

We all know that landing page optimization and creative testing can help boost your conversion rates. 


Still, one area of creative optimization gets overlooked way too often – photo analysis. When it comes to your product or service photos, you have three areas that you need to focus on for maximizing conversions: the photo itself, the product or service within the photo, and the offer presented in the image.


Why are you analyzing your images?

If you’re marketing a product, whether a physical good or service, chances are you’re already thinking about what it will take to convince someone to purchase your product. 

In other words, you’re probably in sales or marketing. And while there are many ways to increase your conversion rate (i.e., making more money), one standard method is adding social proof—the idea that if others have made similar purchases, then maybe I should too. 

But there’s more to it than just showing an image of someone using or enjoying your product; if you want to increase sales and conversions, it’s crucial that your photo selection process be data-driven.

That means gathering hard facts on photo choices and how they affect outcomes—aka analytics.

 For example, which colors work best with what colors? 

Are specific photos better at representing features?

 How can these changes affect click-throughs and purchases?

 The answer to these questions may not seem straightforward at first glance but don’t worry! What follows is a list of considerations to help get you started… you’ll soon see why creative analytics is so powerful for driving business results. 

Let’s start by examining emotion: does emotion have an impact on conversion rates? Yes! 

Emotion makes things easier to remember.

 It’s no wonder sites like Pinterest are so popular—it helps users catalog what interests them without worrying about having a shopping cart handy. There’s nothing wrong with selling products through Pinterest…unless you can sell something cheaper somewhere else, right? 

With social media becoming ever more prominent, brands aren’t left with much room to market their products. If we do decide to share our interest in a product online, we expect interaction from brands.

 And as any user knows well enough by now, interaction doesn’t always come from sellers themselves… After seeing ads that play off our emotions over and over again, we begin to rely on those emotions when deciding whether or not something is worth buying.

 Facebook advertising offers brands lots of opportunities for ad creation: however sometimes keeping things simple provides the most value. 

When you look at your analytics dashboard for Facebook ads are there specific images that stand out as outperforming others? Great! Now all you need to do is replicate that success across multiple images…because even if your analytic dashboard only tells you that image A outperforms image B, it still might be performing worse than images C and D combined. 

Use Case: Groupon – Believe it or not, Groupon didn’t exist until 2009. By 2014 Groupon had grown its customer base beyond 48 million customers worldwide…however achieving such rapid growth has been anything but easy. 

Every new customer represents thousands of dollars in potential revenue yet retaining them remains a huge challenge given customer churn rate hovers around 25%. Perhaps unsurprisingly getting people hooked on discounts wasn’t always Groupon’s main goal – instead, they originally were focused on building brand awareness based around their email lists.

 Groupon’s headlines were also originally less playful and more informative.

 However, it wasn’t long before they realized that simply sending weekly deals to their email lists wasn’t going to cut it—so they decided to explore creative analytics a bit further… Groupon went back to analyzing their Facebook ads and found some interesting things.

 Turns out that headlines with numbers always outperformed headlines with dollar signs, exclamation points, or question marks. Groupon’s copywriters also began to experiment with longer headlines and descriptive tags… Groupon continues to tweak their ads as they learn more about what works best.

 Some of these tweaks have been very successful and some have not—but that’s okay! The point is that creative analytics allow Groupon to make informed decisions rather than relying on hunches and hopes. 

In fact, there are a ton of resources out there for anyone who wants to experiment with creative analytics or social media ads in general.


How many times have you analyzed an image?

There are a lot of companies out there that analyze people’s analytics, especially social media analytics. They try to take all of their customer data and figure out what they should be doing in order to improve their business. 

This is a method commonly referred to as predictive analytics. It’s predicting where your business is going based on past events and actions. 

But what about measuring where you stand now? 

What about looking at images within your posts or website pages?

 Are those images converting or not? 

You could have thousands of likes, shares, retweets and comments but if none of those shares lead to conversions then maybe it’s time you start digging deeper. Just how many times do people look at an image without ever tapping or clicking on it?

 How often do they tap on an image without converting further? 

Why is that happening, and can we fix it?

 Creative analytics takes into account everything possible when determining whether someone has converted after viewing something. The more information we have about why someone didn’t convert, we can adjust accordingly. For example: If users aren’t tapping or clicking even though they seem to like what they see…are your CTAs good enough? 

Maybe you need to make sure that users know exactly why they should tap right away instead of them searching around some more first. Or perhaps your product isn’t ready yet. In that case, will providing users with some other sort of compensation entice them to join your email list in exchange for free shipping once it comes?

 Your creative analytics can tell you these things and much more. Don’t waste valuable time creating content only to have customers leave from sheer boredom and frustration because they don’t know what’s going on–or worse–they don’t know what to do next!


What Can We Learn From Photo Analysis Data?

This should be a fairly straightforward question: what can we learn from photo analysis data?

 As you might have guessed, it depends on what kind of business you’re in. Even if you don’t run an online retail operation, it might still make sense to analyze photos of your merchandise with Convpho before you decide to produce more stock (or discontinue it).

 By analyzing these photos and comparing them to data from other customer sources, retailers can gain important insights into their customers’ preferences and purchasing habits.

 Photo analysis is more than just a way to figure out which images people like best; it helps businesses build connections with customers and better understand how those customers use their products or services.

 Analyzing photos doesn’t take much time—particularly when you use a service that analyzes multiple photos at once—and it provides valuable insight into customer preferences. The only cost is a little bit of your time: a few hours per photo analysis project (or with our Agency Plan, a few minutes on our dashboard)

That seems like small price to pay for any additional insight into consumer behavior! You might not think that something as simple as taking photos or watching TV could teach you anything new about consumer behavior, but there are plenty of useful things you can learn by examining visual content produced by your brand. 

All that viral content may actually help increase sales, so it pays to look closely at photos and videos! What does photo analysis involve?

 Although Convpho automatically analyzes photos according to certain specifications, you do need to provide certain information about each photo before it starts analyzing. 

Here are some factors that Convpho takes into account: Content quality —

How clear is the photo’s text or message? 

How well-lit and focused is it? 

How high-quality is its resolution (i.e., how big can you print it)?

 Using professional equipment can greatly improve photo quality. However, using low-resolution photos or photos taken with amateur equipment won’t necessarily reduce results because convolutional neural networks used by Convpho seem to work quite well even when processing lower-quality photos. Users also tend to prefer User Generated Content where the quality is average. If you’re trying to evaluate photos taken on mobile devices, try cropping off unimportant parts of each photo before uploading it. These cropped photos will give you a more accurate reading of what matters most to viewers. 


Creative analytics tools can help online marketers understand their web traffic and make better decisions. 

Creative analytics programs generally focus on a handful of conversion and behavior metrics, including engagement, session duration, click-through rate (CTR), pageviews per visit and conversion rate. Once you’ve collected data about your users via creative analytics software, you can use that information to determine whether you need to make any changes or if things are working well. 

For example, it might reveal that a majority of your users don’t click on an ad or share it with friends—or maybe they do click on ads but don’t convert afterwards. Knowing what works and what doesn’t is valuable as you fine-tune your campaigns over time. After all, you don’t want to stop measuring or change something based on misleading data. 

Instead, think of creative analytics as a means to measure which parts of your ad campaign work best so that you know how to tweak them in real time. This will ensure more people see your content, get interested in what you have to offer and ultimately buy from you—which is exactly what companies want when they run an advertising campaign! 

In addition, modern creative analytics platforms come with features like heatmaps that allow marketers to visually see how visitors interact with different sections of an image—and where they tend not engage at all. What’s more, these systems also monitor visitor behavior across multiple websites and campaigns. When combined with audience segmentation, these two features are invaluable resources that enable you to pinpoint weaknesses in your website design and tailor future ads accordingly.

 Considering nearly 75 percent of internet users conduct research on mobile devices before purchasing offline, there has never been a greater importance placed on knowing what consumers look at, how long they spend looking at it and when you lose them. Thus, making sure everyone is doing everything they can to drive conversions should be top priority—but doing so requires creative analytics software because no one knows everything about marketing!