EP2: A Data-Driven Approach to Profitable Amazon Keywords

In this second episode, Othmane and Raphaël dive deeper on the Amazon SEO topic by discussing how to generate profitable keywords ideas to improve your Amazon listing, generate more organic traffic, and optimize your margins with a data-driven approach.

Raphael: At the end of the day, what’s Amazon’s main currency, resource, or assets, you may ask? Well, it’s consumer attention. I think that there is no need to reinvent the wheel sometimes again, what’s working for your computer, as you just said, can tell you a lot about what may work for you. 

Raphael: Hey, what’s up, everybody? This is Rafael from DataHawk and welcome to the second episode of the DataHawk eCommerce podcast. In our first episode, we covered the peculiarities of search engine optimization on Amazon, as well as why having a strong, elaborate SEO strategy is paramount for your success.

Today, we will dive deeper into the Amazon SEO topic by discussing how to generate profitable keywords, ideas to improve your Amazon listing, generate more organic traffic, and optimize your margins with a data-driven approach. I’m asking you, have you ever wondered why some of your products don’t get enough organic traffic despite every effort you make to improve your product listings in terms of title, description, bullet points, images, videos, – plus content, amount, and quality of reviews, or even advertising efforts and even sales velocity? It’s probably because you are missing the data-driven part. Improving your listings without analyzing the competitive dynamics in your market and understanding Amazon’s ranking algorithm can cost you a lot.

Now, before getting started, and if you’ve never heard of DataHawk yet, we are a turnkey software analytics and recommendation platform that helps brands, retailers, and agencies increase their sales, optimize margins, boost productivity and gain insights into selling on Amazon. And I’m here today with Othmane Sghir, CEO and co-founder of DataHawk. Bonjour, Othmane.

Othmane: Bonjour, Raphael.

Raphael: Maybe we could start our discussion by covering an overview of the different key points that affect Amazon’s sophisticated ranking algorithm, and go into the making of profitable keyword research and SEO strategy. 

Othmane: Definitely. I would say that they fall within two distinct buckets. The first bucket relates to item data quality and comprises text and image-based data such as the title, description, bullet points, back-end search terms, images, videos, and a-plus or enhanced brand content. The second bucket relates to comparative data points and comprises numerical data including price, reviews, ratings, sales velocity, and sales or best sellers rank, conversion rates, predicted rates, return rates, seller performance, and maybe time depreciates as well. 

Raphael: All right. Well, the first bucket around item data quality sounds straightforward and extremely logical. In some ways, data quality represents the basics of SEO on Amazon. The second bucket around what you call comparative data points is interesting and isn’t something we hear about often. So, could you elaborate more on that first, especially on your use of the word comparative in it?

Othmane: In our discussions here at DataHawk with brands, we’re surprised how many professionals are oblivious of the importance of comparative data points in SEO on Amazon. They’re often surprised when we showed them concrete examples of debt based on data points we collect. So, my use of the word comparative hints at the fact that the numerical data points related to a product are part of a broader ecosystem: a large digital shelf on Amazon where other competing products are fighting for consumer attention.

So, when Amazon’s ranking algorithm looks at your product, it compares its price to how other competing products that are fighting for similar search queries are priced. It compares its conversion rates, it compares how well-rated it is, and so forth. At the end of the day, what’s Amazon’s main currency, resource, or asset, you may ask?

Well, it’s consumer attention. Ultimately, the algorithm wants to make this limited pool of attention and time spent on the website as profitable as possible, while also maximizing consumer satisfaction. And the numerical specificities of your product in terms of price, reviews, rating, sales velocity, conversion rates, CTRs, return rates, seller performance, time depreciates, and final basket size are definitely affecting that. 

Raphael: To better illustrate this point using data, let’s play a game. So, I’ll give you a few keywords or search queries pertaining to different products or categories, and you look them up on DataHawk to share with me insights based on this. All right, so here’s my first keyword. Memory foam mattresses. All right. Memory foam mattress. 

Othmane: I’m crazy about memory foam mattresses right now. You’re looking for a mattress? 

Raphael: Yes, exactly. No, for real.

Othmane: You’re not sleeping well. I can see that. So let me look at that. Since, like the highest price product ranking organically on page one on Amazon.com for the search query queen-sized mattress is priced at $291, versus $369 on page two, and $520 on page three. On a median basis, products ranking organically on page one are priced at $175, have around 1,200 reviews, and a rating of four stars. So, I hope I haven’t lost you there now. 

Raphael: No.

Othmane: And I just figured out that I answered this for a queen-sized mattress instead of the memory foam mattress that you asked for. 

Raphael: That’s all right. I’m also looking for, a queen-sized. 

Othmane: So, in other words, based on the data I’ve just shared, a Casper or Tuft and Needle mattress that’s usually priced north of $600 would never make it to page one organically for the search query queen-sized mattress on Amazon dot com.

Raphael: Organically, huh?

Othmane: Yeah, organically. And that’s simply because the competitive dynamics in terms of competing products that are relevant for that same query make it so. And it also means that it is extremely competitive and hard to rank on page one for that search query, as you would need to generate so many sales and reviews while having a super high-quality product. So, looking at the data, actually, 75% of the products ranking organically on page one for queen-sized mattresses have more than 480 views. 

Raphael: That’s interesting. So, what you’re saying here is that there is some sort of statistical scorecard or profile for every single keyword or search query on Amazon and that your product needs to fit in in order to appear organically in the search results.

Othmane: Exactly, and this is the sort of analysis that our software DataHawk makes easier to access for our customers. We notably compute competition scores related to keywords by looking, for instance, at how many reviews and ratings you would need to rank on page one organically. 

Raphael: Right. So, through the first example, we mainly covered price, ratings, reviews as being critical to organic rankings. In my second keyword or search query suggestion, maybe could we cover the sales velocity part along with other aspects you mentioned? Let’s do this for the keyword running shoes. Shopping around for running shoes. Exactly. Okay. And we need to race, better sleep, and better health. 

Othmane: Good for you. So, sure. So, when you look at running shoes in the US, again, on Amazon.com, Asics is killing it there with a strong share of voice for it. They’re actually dominating the organic rankings with their Men’s Gel-Venture Six running shoe that has consistently ranked first for that keyword over the past 30 days alone, for instance.

The product, when you look at it, fits all what we previously discussed and which helps it rank for running shoes. Being priced at around $50 has 4.6 stars with over 8000 ratings, which is comparatively amazing and perfect when looking at other products it competes with for the same search query. Now, what’s also interesting is the fact that it has also consistently ranked first in the trail running and road running browse notes or categories on Amazon in terms of its best-sellers rank or sales rank. Meaning, it has consistently been the top-selling product in those categories. As to the third-party sellers that have been distributing, sorry, the product and winning the buy box the most, they all have a set of feedback, I can see here, in the 90% to 99% range over the past twelve months.

Raphael: Right. So, in other words, the product is also having a certain sales velocity that’s necessary to rank on the top of the fold for a highly competitive short-term search query like running shoes, further helped by a top-notch distribution. 

Othmane: Yeah.

Raphael: So now to summarize what we discussed, on top of nailing the perfect copy in terms of the product listing page, one is also to pay attention to how the product fits within the competitive dynamics of the keyword it’s trying to rank for. Notably, in terms of pricing. Othmane, are there any other data points one has to look at in order to know which keywords to focus on? 

Othmane: I would say probably the estimated monthly search volume tied to each keyword in comparison with the competition level on that same keyword as reflected, for instance, by the number of products indexed on the search results on Amazon for it. Ideally, you want to optimize for keywords that have a high consumer demand relative to the number of products available for purchase for it. Right? And the other thing would be the advertising environment around those keywords. How many sponsored results are there over time in the first and second pages relative to the total results? How often? Are bids high or low? Is there a handful of brands that are dominating advertising on those search queries? How big a budget do they probably have? Those kinds of questions. 

Raphael: Right. One area that we overlooked in our discussion is the first bucket around item data quality. How can one be data-driven on that front and build the perfect listing copy? 

Othmane: Well, this can be achieved mainly by having the mindset of running AB tests and doing constant optimizations in a product listing in terms of the keywords used in its title, description, bullet points, and back-end search terms, on top of the images, videos and a-plus content being used. Using an organically ranked tracker, such as the one provided by our software platform here at Datahawk, can help measure and assess the impact of each key iteration by looking at how it would result in changes in organic rankings for the keywords that are being optimized for.

Also, I always recommend running benchmarks against computers because it’s simply always helpful to see what’s working for them and compare your performance against theirs. 

Raphael: Yeah, if it’s working for them, it will work for you, I guess. 

Othmane: Definitely. Yeah.

Raphael: So, a lot of what we discussed kind of requires already having a set of keywords one wants to optimize for and analyze. Now, what are the tools that are available for brands and sellers to actually generate keywords ideas which they can then go on to analyze further to see if they’re interesting, profitable, and relevant for the product, as we previously discussed? 

Othmane: I think that there is no need to reinvent the wheel sometimes again. and what’s working for your competitors, as you just said, can tell you a lot about what may work for you now. So, the first approach I’d recommend is looking at the top-performing and top-selling products you compete with and monitoring and analyzing them.

Look at what keywords they are using in their copy, and so forth. One tool that’s available on Data Hawk again, and which can come very handy on that front, is our keyword lookup tool that allows you to unveil keywords for which any product on Amazon has ranked organically at some point in time. So, I would definitely recommend using this tool on the best products to compete with to extract keywords ideas.

We also actually provide an AI-based keyword generation engine that can also be helpful, and for both of those tools, you can basically run on any product on Amazon. You simply put DSC in there and run the analysis and you extract the insights.

Raphael: Your product and your competitor’s product. 

Othmane: And your competitors as well. No need to connect any sort of seller or vendor central account. 

Raphael: What about advertising data? Any way this can be leveraged as a good source of keywords generation? 

Othmane: Yeah, definitely. Automated keyword bidding handled by Amazon is always a great thing to run first for a few days on products in order to generate data points, crunch them, and then pick high converting keywords with a great return on ad spend. Then the second and most helpful step is definitely running phrase match bits on short (()) search queries of let’s say two to three words and leveraging the customer search term report to surface exact search queries used by customers to purchase products. Then, similarly, picking those that are most profitable.

Raphael: Yeah. Makes sense. Then I assume that taking those keywords and monitoring and analyzing them on that can help narrow them down further and even help better understand how bidding on those keywords and generating paid sales can affect organic rankings and can be smartly used as a way to pilot a product’s sales velocity. All right, well let’s not get too excited in here, and let’s keep that for another time, another episode. To wrap up our discussion, Othmane, do you have, uh, short advice for our listeners? 

Othmane: Remember the two buckets we discussed. Item or listing data quality and comparative data points. Don’t stop at simply nailing the basics in your listing copy. Be data-driven on optimizing your listings to the best of your ability. And, importantly have a comparative and market-driven approach on your keyword research strategy. Remember that there are critical competitive dynamics around each keyword, so, you need to analyze those. Be cognizant of their importance and build your listing and your organic and paid search strategies around them and DataHawk can help you do that and much more actually.

Raphael: Awesome. Thank you, Othmane, and thank you, everyone, for listening to our second episode. We hope we provide you, with some good juice on this DataHawk eCommerce podcast, and again hope to see you soon. Bye-bye.

Othmane: Bye.