July 2023 Prime Day Recap

R.J. Lehman
Intent & Effect
Published in
5 min readJul 13, 2023

As Prime Day drew to a close, the team at Bullwhip took a quick look at the publishers that drove the most engagement throughout the 2-day event. Below we summarize some of the more interesting and salient insights we came across.

Photo by ANIRUDH on Unsplash

And there it is — another whirlwind Prime Day wraps. For the team at Bullwhip though, this marks a special occasion as the first Prime Day for which our data systems were up and running, as well as the inception of Bullwhip’s newsletter, Intent & Effect, which will take a data-driven approach to elucidating trends and opportunities in the commerce content / intent media space.

So what did we learn? Here are the quick hits:

1. Top Stories continue to gain importance.

As most publishers are now aware, continuing to diversify audience acquisition away from traditional organic search is critical. This can mean leaning into non-search sources (recirc, email, social, network and/or paid), but let’s not forget the other gems within search (ahem, Google) itself, namely Top Stories, Answered Questions and Discover.

Focusing on the first two (as public Discover data is limited), we can actually look at Top Stories vs. traditional organic “SERP Links” in an apples-to-apples fashion based on page rank / position. We did exactly that, every hour, across the leading 600 Prime Day related keywords, and were surprised to see Top Stories taking up a massive 32% of the combined total.

We dive into this a bit later to provide more context on how to think above top stories (as well as which ones crushed it this Prime Day), but first, let’s address traditional search…

2. Amazon earns a meaty share of search share of voice, but publisher presence still leads.

We were pleased to see publishers continuing to have strong presence in search across the top Prime Day keywords — judging by historical page rank click-through rates (which won’t entirely hold for navigational traffic), we estimate the share of voice (SoV) of the top 100 publisher domains to be 75% of the total. Amazon.com is by far the largest single domain with 12% SoV, and their white-label content site aboutamazon.com garnered another 3.5%.

For better or worse, the intensity of search audience acquisition during Prime Day causes this to be a “winner take most” game, with smaller domains collectively only achieving 3 p.p. less than amazon.com alone.

3. A number of stalwart commerce domains continue to perform well, while newer commerce entrants gain ground.

A number of intent media leaders we’re accustomed to seeing at the top of the pack led in the Top 30 by hour during July 12, 2023 (chart below). These include CNET, NY Mag’s The Strategist and Vox Media sibling brands The Verge and Polygon, Forbes Vetted, Business Insider, NYT’s Wirecutter, Tom’s Guide, USA Today’s Reviewed and Good Housekeeping.

The below chart shows hourly performance (July 12, UTC hours) of each of the top 30 domains, where the thickness of the line represents that brands estimated share of voice during that hour, based on their position across the 600 keywords analyzed during each hour.

At the same time, some media brands that are newer to commerce content also stood out, namely NY Post (whose 108 Best Deals was the highest exposure article during July 12, second only to Amazon.com’s landing page), Mashable, ZDNet and others.

The rank ordered list of the top performance articles over the course of the 2nd day is below.

Top Performing SERP Links

4. A slightly different group of media brands won the day in Top Stories traction.

Although Google New’s “Top Stories” block, which as noted is gaining presence in organic results, is open to all publishers of relevant, prominent, authoritative and fresh content, it’s perhaps not surprising that news brands and portals like cnn.com, nytimes.com, nbcnews.com, yahoo.com show up higher on the list here.

CNN Underscored’s 151 Best Amazon Prime Day Deals 2023 led the pack in Top Stories distribution, and the other leaders are below.

5. Recency and update frequency are likely a key differentiator for Top Stories performance during Prime Day.

On average, the article in position 1 of a Top Story block was updated every ~13 hours, 70% of these articles had been updated between 0 and 24 hours, and now article in position 1 (across ~9,000 top stories blocks analyzed) was more than 72 hours old. Not much of a surprise but confirmation of the hustle that differentiates the leading publishers during Prime Day.

Moreover, as shown in the chart below, the average hours since update for positions 2–4 was not much higher, although the spread was a bit wider. The more recent publish time averages for positions 5 and 6 may reflect that Google is only populating this many stories when there is an abundance of high-quality and recent content (position 6 had the most recent average publish time at 9.9 hours, but was only present in ~38% of the Top Story blocks we reviewed).

Looking at average recency (or age) of Top Stories by hour of day (again, in UTC), we can see that overall freshness increases as the day went on, which makes sense given the active “news desk” approach most leading publishers take in running their Prime Day playbooks.

6. People are curious about all sorts of Prime Day things.

Just for fun, and to satisfy our own curiosity, we looked into what questions appeared most frequently in the “related questions” block. They are below.

And although we can’t get as granular data about like traffic / engagement volume from these Q&A blocks, we can share a bit of context around which brands and domains are showing up with answers most frequently. Kudos to CN Traveler, ZDNet, NBC News, CBS News, People, Good Housekeeping and others.

Concluding Notes

If you’ve read this far, many thanks for it! Moreover, we’d love any feedback or follow up questions you have. Ping me directly (email in bio), or the team at partner@bullwhip.io.

The supporting data for the above analyses is in this Google Sheet. Feel free to reuse any of that or this article with attribution as described on the cover sheet of the workbook.

Looking forward to Early Access in October!

--

--