Tracking Honey — How To Measure the Impact of Automated Couponing on eCommerce

Digital couponing is widely used among online shops. And rightfully so. Promotional discounts can be tools for user-attraction, incentivization, retargeting, upsell, or churn prevention.

Since there are so many promo codes floating around the WWW, some of which might be targeted campaigns, it is impossible for an online customer to be aware of all ongoing promotions and whether there are available discount codes that could save them some serious money on their order. This is where tools like Honey join the party.

Honey offers a popular browser extension, which is triggered while a user is completing the checkout funnel in one of the 40,000+ supported online stores. It automatically searches, displays, and applies the best available coupon code at checkout. It is a handy tool for online shoppers and a lovely concept.

Shop owners, however, might not always be aware of this background activity during checkout. One major downside of extensions like Honey influencing their conversions is the misinterpretation of the resulting sales data. As an analyst, one might be inclined to falsely attribute the usage of a certain coupon code to a specific marketing campaign. However, the majority of those conversions could have simply used Honey to apply that code, while never even having been reached by that campaign. On top of that, the strategic purpose of couponing seems to be defeated in the Honey-case. Coupons are meant to attract or incentivize potential customers. But since users are already at the very last step of the checkout process when Honey starts doing its magic, they might be likely to make that purchase anyway. Regardless of whether they get that promotional price or not. Hence, giving away those discounts right before conversion could very well be considered losing money. Of course, that is just one way to look at it. Others might argue that the prospect of a discount could give users the final nudge to convert or even trigger a larger purchase.

I am neither advocating nor opposing such browser extensions here. Tools like Honey are incredibly user-friendly and even cooperate with a lot of online shops. Working with analytics, my simple goal is to use some nifty analytics to give eCommerce sites the possibility to measure the impact of automated couponing on their sales. The resulting data will give them a clearer picture of their transactions as well as valuable insights that could potentially be used to design innovative partnership models with such extensions.

How to Track The Impact of Honey on Sales Data

Step 1
Use your analytics solution to capture applied coupons and discounts for each transaction. This can be done via custom events or with built-in methods, like the promo code tracking in Sales Insights by AT Internet.

Step 2
Implement a script that detects whether Honey is active on your site during a session. If it is → Trigger a tracking-event.

Step 3
Filter and segment your sales data on conversions, during which the Honey-event occurred. Visit-based segmentation helps you single out converting sessions, in which Honey was active. User-based segments can help you build cohorts of Honey-users.

Implementation & Code

Here is some background on the implementation and the full tracking code:

• If Honey is active during a checkout process and it has found applicable coupons, it displays a pop-up to notify the user:

• In order to do so, the extension is appending a <div> element to the site’s code with the ID ‘honeyContainer’.
• This is the activity that I want to detect. However, I cannot simply check for that div once a page is loaded, because Honey is not yet active on page load. The extension can be triggered at any point during the checkout. That is why I need to constantly check whether the DOM is being modified and whether this particular div was part of the modification. Luckily, the MutationObserver interface lets me continuously watch for modifications of the DOM tree.
• If at any modification-event, the Honey-div is being inserted, I can detect that and action on it. In my particular case, I am sending a custom event called ‘Honey active’ to my digital analytics solution AT Internet, where I have also configured my eCommerce tracking.
• In order to limit the impact of the script on page performance, I am stopping it, as soon as the element has been found.

Here is the full code — jQuery required:

Full disclosure: The script above will only check if Honey has been actively displayed on the site during checkout. This does not guarantee that the user actually took the last step and applied the coupons suggested by Honey.
In my opinion, however, if a purchase has been made with a valid discount code and Honey has been used in the same session, one can assume that the majority of those discount codes have been applied by the extension.

Reporting on Honey

That’s it! You can now explore the impact of extensions like Honey on your analytics data.

→ What is the share of turnover that stems from Honey-conversions?
→ Do Honey-conversions have a higher average cart value?
→ Does Honey incentivize first-time purchases?
→ What is the exact amount of discount that has been granted to customers due to Honey (‘lost revenue’)?
→ How does automated couponing compare to the overall usage of discount codes?
→ Does the presence of Honey impact the drop-off points during the checkout funnel?
→ Do visitors who benefit from discounted prices via Honey have a higher retention rate?

Bottom Line: We can’t all be Winnie the Pooh. For eCommerce sites, Honey can be a double-edged sword. Time to use the power of digital analytics to add a layer of transparency.

— If you have questions about the code or if you’re simply up for talking analytics: Feel free to reach out. Shout-out to Alex Louden for his article on the MutationObserver API —