Helium 10 Cerebro vs. WordTree

As much as I encourage you to read this entire blog post about Helium 10 and WordTree, I understand that your time is important, so for those that don’t want to read the details, here’s a table demonstrating what we found in our comparison of WordTree’s reverse ASIN tools to Helium 10 reverse ASIN tools (cerebro).

We purchased 2 reverse ASIN reports from Helium 10’s Cerebro so we could compare them to our WordTree reverse ASIN reports. We used a garlic press and some marshmallow roasting sticks as our demo products of choice.  Here are the results we found.

Garlic Press
WordTree Helium 10 Cerebro
Initial 403 Initial 445
Keywords Removed Keywords Removed
Duplicates Duplicates
Questionable 66 Questionable 170
Stems Stems 82
Spelling Spelling 6
Remaining 337 Remaining 187
% relevant 83.62% % relevant 42.02%
Jungle Stix
WordTree Helium 10 Cerebro
Initial 626 Initial 1052
Keywords Removed Keywords Removed
Duplicates Duplicates
Questionable 94 Questionable 278
Stems Stems 355
Spelling Spelling  10
Remaining 532 Remaining 409
% relevant 84.98% % relevant 38.87%

As you can see from the table. WordTree had an average total keyword phrase accuracy of 84%, while  Helium 10 Cerebro had an average total keyword phrase accuracy of 40%. Also, WordTree found an additional 273 keywords over Cerebro, making WordTree the obvious winner.


The Details

It’s a question we at WordTree get a lot – What separates you from Helium 10’s Cerebro or other reverse ASIN tools?

We can sit here and tell you “we just are,” or, “because WordTree was written by search engine experts,” but really, does that really mean anything to you – the Amazon seller?

So, today, we’ve decided to put our money where our mouth is.  We will be directly comparing results from WordTree vs. results from Cerebro.

For our comparison of the tools, we will be picking two ASINs that most Amazon private labelers/sellers are familiar with.  The Jungle Stix product that was launched by our friends over at Jungle Scout, and a Garlic Press, which has been popularized as the go-to example product by Scott Voelker in his podcasts over at The Amazing Seller.  So, without further ado, lets jump in!

Jungle Stix - Our First Example Product for Testing Helium 10
Bamboo Marshmallow S’mores Roasting Sticks by Jungle Stix (ASIN: B017VXKVXE)

 

Garlic Press - Our Second Example Product for Testing Helium 10
ORBLUE Propresser Stainless Steel Kitchen Garlic Press (ASIN: B00HHLNRVE)

Getting Our Data

We started by getting our reports over at Helium 10’s Cerebro.  We put in the two ASINs B00HHLNRVE and B017VXKVXE and downloaded our results.

We got our data over at WordTree as well.  We purchased the Jungle Stix keyword report by using ASIN B017VXKVXE with competitors B01M725DNP, B01MR827H3, and B0188V4N0A.  We then bought the garlic press keyword report by using ASIN B00HHLNRVE with competitors B01M1LLHFK, B00I937QEI, and B00HEZ888K.

The Results Are In!

Cerebro

For the Jungle Stix report from Cerebro, we got back 1,052 keyword phrases!  That sure is a lot of keywords.  However, that doesn’t tell the whole story, we could still be getting a lot of junk that isn’t useful.

For the garlic press, Cerebro found us 445 keyword phrases!    Hopefully, this provides us a decent list to get our PPC and Amazon listings optimized for the searches that Amazon customers are typing.

WordTree

For the Jungle Stix keyword report from WordTree, we got back 625 keyword phrases.  Ouch, not as many as Cerebro!  But sometimes, accuracy is more important than volume.

For the garlic press, WordTree found 403 keyword phrases.  Also not as much as Cerebro!  Hopefully, though, this provides us with enough information to build a reliable listing and a robust PPC campaign.

Cerebro Duplicates

Looks like Cerebro is on top of their duplicates!  For both Jungle Stix and the garlic press, there were 0 duplicate keyword phrases found.

WordTree Duplicates

WordTree also has returned 0 duplicates for both Jungle Stix and the garlic press – so everything is already unique and filtered!  Sweet!

Questionable Keywords

For the next comparison, we are going to take a look at the “questionable phrases”.  What exactly does that mean?  Well, for this test, we are defining a “questionable” phrase as a keyword phrase that most likely doesn’t have the potential to earn a sale from a customer typing it in the search bar.  Keywords that don’t fit this quota are deemed as “questionable phrases”

Cerebro – Questionable Keywords

After filtering the duplicates, we are left with 1,052 unique keyword phrases for Jungle Stix and 445 unique phrases for the garlic press.

After reading through every single result on the Jungle Stix report by Cerebro, we found that 278 of the keyword phrases were questionable.

What do we mean by that?

We’ll let the results speak for themselves… Below are some questionable phrases returned from Cerebro…

jungle snugs hooded towel
break the safe
six days of creation
living the life that you are
rest warmer

This list goes on for 278 keywords.  That was a lot of manual filtering, and it took us an extremely long (and tedious) time to remove them all from our list.  I really hope I don’t have to do that ever again…

What about the garlic press?  Well, after reading through every result in the garlic press report by Cerebro, we found that 170 of the keyword phrases were questionable.  

Again, let me show you some examples…

easy hang up
handle reduce
produce storage for kitchen
home discount kitchen
images in motion

So, overall, 30% of Cerebro phrases were questionable, and in our opinion, wouldn’t result in sales from customers typing them.  It makes you wonder – how are they finding their keywords?  Where did all of those garbage keywords come from – and why does their system think they are relevant?

Beats me… but let’s take a look at the results from WordTree now.

WordTree – Questionable Phrases

For Jungle Stix, we have 625 possible keyword phrases in our keyword research report.  But, just how many of those are questionable?

After reading through every single result in the Jungle Stix report by WordTree, we found that 15% of the keywords were questionable.

Here are some examples of words we found that probably aren’t bringing sales…

campfire fun
heavy duty fire pit
huge marshmallows
the perfect campfire grill
jungle fire

For the garlic press, we have 402 possible keyword phrases returned by WordTree.  So, how many of those are questionable?

After reading through every single result in the garlic press report by WordTree, we found that 16% of the keywords were questionable.

Again, some examples that probably aren’t bringing sales and could be safely ignored for PPC/listing optimization…

cleaner for garlic press
best peeler
steel hand press
point press
cooking press tool

So, in the end, about 16% of the results from WordTree were questionable and could (most likely) be safely ignored.  It’s better, but still not perfect, but let’s take a deep even further into the results provided by Cerebro vs. WordTree

Keyword Stemming

Keyword stemming is taking the “base” of keywords and filtering out the ones that don’t matter because they are interpreted the same by Amazon’s search engine.  For example, in a search engine, like Amazon’s product search, whether the customer types:

garlic press
garlic presses
garlics press
garlics presses

They are treated the same when it comes to optimizing your product listing or building a PPC campaign.  In this example, all you need is “garlic” and “press,” and you’ll index for all of these search queries.  Why?  Because they both stem to the following in the backend:

garlic press -> garlic press
garlic presses -> garlic press
garlics press -> garlic press
garlics presses -> garlic press

Make sense? If not, here’s another example:

Fishing Rods -> Fish Rod
Fish Rods -> Fish Rod
Fishes Rod -> Fish Rod
Fished Rod -> Fish Rod

When you put these into your listing or PPC, you only need to include one version, and Amazon will index you and/or advertise for all of them.  So, essentially, any time these phrases occur as a keyword phrase in your campaign or your listing, Amazon treats them the same, and you are wasting possible words to index on your listing if you include each and every variation.

That’s an introductory overview of keyword stemming; we could write an entire blog post about it, to be honest.  Continuing, let’s take a look at the two keyword tools in this regard, and see what’s happening between them.

Cerebro – Keyword Stemming

After we are left with the unique keywords from Cerebro, how many of those phrases are duplicated by not performing keyword stemming on our phrases?

After taking out duplicate phrases, we are left with 774 unique phrases for Jungle Stix, and 275 for the garlic press.  Let’s see what happens to those lists when we remove duplicated keyword stems.

Jungle Stix – 774 -> 419 remain after keyword stemming (46% Loss)

Garlic Press – 275 -> 193 remain after keyword stemming (30% Loss)

So after we’ve de-duped, we are still only able to use an average of 58% of the remaining keyword phrases from Cerebro.

WordTree – Keyword Stemming

WordTree already takes care of word stemming for you, so we are still left with the same number of results.  Both reports have already been normalized and stemmed, so we still have 532 right phrases for Jungle Stix and 337 useful phrases for the garlic press!

Amazon’s Spelling Corrections

An often overlooked part of keyword research for Amazon spelling auto-correction system.  What does this mean?  Well, say you misspell the word “marshmallow” as “marshmellow,” Amazon will auto-correct your search, making the search for “marshmallow” instead.  Take a look for yourself:

helium-10-not-accounting-for-mis

What does that mean for PPC and keyword optimization?  Well, in short, including the word marshmellow or phrases with it is useless. Since Amazon is auto-correcting these mistakes, there is no reason to correct them ourselves.

Lets again analyze each tool to see if these are being taken care of.

Cerebro – Spelling Corrections

For Jungle Stix and the garlic press, we are left with 419 and 193 phrases respectively.  How many of those are automatically changed to something else by Amazon’s search engine due to spelling errors?

Jungle Stix -> 419 -> 10 Auto-Corrected -> 409 remain (2% Misspelled)

Garlic Press -> 193 -> 6 Auto-Corrected -> 187 remain (3% Misspelled)

WordTree – Spelling Corrections

WordTree already takes care of spelling corrections from Amazon’s A9 search engine, so there is nothing to do at this step.  All of the remaining 532 / 337 phrases are still good to go!

Conclusion

Using Cerebro, we started with 1,052 phrases for Jungle Stix and 445 phrases for the garlic press.  By the time we removed duplicates, questionable phrases, stems, and spelling corrections, we were only left with 409 and 187 phrases respectively.

This gives us a loss of 60% of our original phrases returned by Cerebro, leaving us with only 40% that are usable for PPC and listing optimization.  Is your time worth sitting there and doing everything I just did to get a clean list of keywords?

Using WordTree, we started with 625 phrases for Jungle Stix and 403 phrases for the garlic press. By the time we removed questionable phrases (as that’s the only thing not automatically and fully removed from the report), we were left with 532 phrases for Jungle Stix and 337 phrases for the garlic press.

This gives us a loss of 16% of our original phrases returned by WordTree, leaving us with 84% of our keyword phrases ready to use in PPC and our listing.

WordTree was written by people who understand search engines.  We strive to offer the most accurate and capable keyword research software on the market… sign-up today to start optimizing your listings and find the best keyword phrases for PPC campaigns.

  CEREBRO WORDTREE
KEYWORD RESEARCH
ONE TIME PAYMENT
REMOVES DUPLICATES
REMOVES STEMS
AUTO-CORRECTIONS (SPELLING)
MINIMAL FILTERING

You can download the reports used for comparison here.

Questions?  Comments?  Write us a message below.

www.wordtree.io

Disclaimer:  WordTree is in no way affiliated with Helium 10.  Any results in this post should be interpreted as the independent opinions of WordTree LLC.