Once you set up your DataHawk Sources and Destinations, you can access various datasets related to keyword suggestions, keyword lookup, and keyword placement.

We have compiled these datasets to provide the data you need to understand your usage of keywords better, enhance your keyword research efforts, and improve your visibility.

Datasets

Connections currently provides three datasets, which are accessible to users who have tracked products.

1. Keyword Suggestions
2. Keyword Lookup
3. Keyword Placement

Overview

All of the data listed below is available at ASIN-level.

1. The Keyword Suggestions dataset provides a range of keyword ideas to help you expand your list and uncover new opportunities. This dataset applies to tracked products and is equipped with several metrics that can help you evaluate the potential of a keyword and decide whether to incorporate it into your SEO strategy.

   • How do we find keyword suggestions?

   Here are the steps that we follow to generate new, relevant keyword suggestions every week for a tracked product:

   1. We identify the closest competitors of the product and extract their category. For more information on how we identify competitors, click here.
   2. We extract a pool of keywords from the competitors’ categories (the ones that are relevant to the product).
   3. We assess the significance of keywords and select only the most relevant ones for the product.
   4. We exclude keywords that mention colors, dimensions, or brands, and include relevant metrics to help you compare and benchmark keywords.
   • How are the related metrics computed?
     • A product's Competitivity Score for a specific keyword evaluates its likelihood of appearing in the search results. The score usually ranges from -100 to 100. 100 means that there's a high chance your product will reach the top ranks because it has much better stats than other products appearing in the keyword, -100 means there's almost 0 chance of seeing your product rank high even if you use the keyword, and 0 is neutral; your product is as competitive as other products. To compute this score, we compare your product's page to the top (three) products that rank for the keyword. We consider factors such as reviews, ratings, prices, and relevance to the keyword.
     • The product's Monthly Reach Potential estimates how many people may see your product when they search for a particular keyword. The higher the score, the more likely people will see your product. To compute this score, we estimate how many people search for that keyword each month and combine it with the competitivity score of your product.

2. The Keyword Lookup dataset provides all keywords for which we have seen your tracked products rank organically in the last seven days. This dataset enables better analysis and discovery of the keywords in which your tracked products are visible.

3. The Keyword Placement dataset provides insights into how you use keywords on your product listings. We analyze the different sections of your listings (title, description, bullet points) daily and display which keywords appear in each of these. This lets you see which keywords you are using and analyze your SEO strategy's effectiveness.

Destinations Datasets

Using destinations, your accompanying DataHawk-powered Snowflake database contains multiple tables about Keyword Research data. These tables contain a very exhaustive list of data points as columns and are organized into schemas. All Keyword research data is available under the REPORTS schema.

These datasets can be used to build dashboards on your favorite BI solution.

To see a complete view of all columns within these tables, refer to our Table & Column Referential.

Considerations

Keyword Suggestions

Keyword Suggestions are for now only available for English-speaking marketplaces (US, GB, CA, AU, SG, IN, AE).

To maintain relevance, this dataset avoids including keywords that mention specific colors, brands, or dimensions. Nevertheless, words with multiple meanings have been retained. For instance, "moccasin" can refer to both a type of shoe and a beige color. Therefore, we advise you to proofread carefully when using these words to ensure accuracy.

Refresh Frequency

Data is updated daily for Keyword Suggestions and Keyword Placement datasets and weekly for the Keyword Lookup dataset. That is to say that if you track by noon UTC, you will have suggested keywords by 6 am UTC.

Why is there no suggestions for my product?

We are able to suggest keywords for 90% of products. Usually, we fail at providing keywords because we could not extract any obvious category and because we never saw your product ranking in any keywords. Hence, we would advise you to either refine the category choice and/or track keywords relevant to your products.

To find keywords, we look at the frequency of appearances of each one of your competitors in every browse node. We then use a pre-computed list of keywords assigned to each browse node to map your competitors to keywords. Then, we use an ML algorithm to sort out these keywords based on semantic similarity. Hence, if we did not find any competitors for your products or if the keywords we found had a too low semantic similarity, you may not find any suggestions. This is because we favoured precision over coverage in order not to overwhelm you with bad suggestions.