In this case, the primary category for Intercom is `User onboarding and Engagement` which can be broken down separately to `User Onboarding` and `In-app Support` or `In-app Chat`. That will suddenly spike up the sanity of both categories (maybe we can even normalize further - siftery has more than 700 categories/tags, a first for any such platform).
Categorisation of a data set like this can be non-trivial and will false positives. We can cherry-pick such a false positive and ignore all the actual positives. That said, thanks for this feedback - this is exactly what is needed to improve status quo. We are on it and will report back in a few hours with the category split I suggested in the previous paragraph. Do let me know if you think it can be broken down into a different type of granularity?
You can submit products through this form - https://siftery.com/submit-product, our team will do a quick review, and push the product through if it's legitimate and live.
We have 40K+ products in the DB, so you are in good company :)
Alternativeto is great. _Siftery_Product_Substitutes_ is an attempt to analyze the migration of one product to another, across companies to begin with, and later across several other segments and industries. Where we don't have any switch data, we are only showing vanilla alternatives (sorted by NPS).
This is merely v1 and we intend to iterate quickly on this dataset and provide more intelligence/decision support for buyers when they are looking for alternatives to switch to for a specific product or across a category.
Search for the top alternatives for over 40k B2B software products.
We’re doing a couple of things here that this community might find interesting:
A) Actually tracking when companies start and stop using a piece of software
B) Using this “switch” data to calculate a probability that the switches are a true substitution and then rank the top substitutes for each product - based on actual switching behavior. We use a weighted average where the switches are weighted according to how much the product’s categories overlap (every product is tagged with 1-5 tags). For example, Intercom and Drift are closely related so when a company stops using Intercom and starts using Drift that’s heavily weighted. However, a percentage of the companies who stop using Intercom and then start using Zendesk are effectively substituting Intercom with Zendesk.
- You can use search to find a product, or start with the ones below:
>"When companies stop using WordPress, its most frequent substitutes are HubSpot Website Platform (32.6% of the time), followed by Drupal (21.9%), and by EPi Ektron CMS (6.9%)" (https://siftery.com/wordpress/alternatives) //
I don't understand why the first listing Craft CMS isn't one of the hot alternatives that people switch to as mentioned in the prose? Why is it even in there? Why is the third hot substitute not even in the list below of alternatives?
What we're doing with Substitutes is to complement the NPS ranking with a mapping of actual switch behavior. It's entirely possible that a lower-rated product is a much more popular substitute for a product if it's a better fit (e.g. targeting SMBs or Enterprise).
An example:
In the Recruiting/ATS space, Greenhouse and Lever are each other’s most significant competitor.
Meanwhile, companies looking to replace iCIMS by far most frequently end up using Oracle Taleo (and vice versa)
We generally see company size and employee count is highly predictive of which products they’ll consider and we can see a progression as companies grow out of services catering to SMBs to more fully-featured alternatives.
Why does it say that the serverless framework is something that runs on top of AWS lambda not a replacement? Or does it mean to use straight AWS lambda without a framework?
I like the idea of the site and I have used similar sites in the past.
A great use case where we can increase the granularity of categorization/tagging and potentially separate the infrastructure provider and the implementation layer on top of it.
0) We analyze products that have a public footprint.
1) We analyze products that don't have a public footprint.
2) For 0), we crawl the internet periodically and create a time series curves of B2B product usage for companies.
3) For both 0) and 1), our verified community has trained our algorithms with tens of thousands of data points on usage and sentiment, thereby reinforcing or nullifying these deductions.
4) If you have stopped using Angular and started using React (or vice-versa now that Angular 5 has better server-side rendering), it is deductible at a global context.
We are looking into the data discrepancies, it might take a bit of time to get through the individual errors. DNS + Registrar data seem to have brought on false positives. We need to beef up on Front-end tech too. The real time mapper needs work.
Humbling comments really, we have our work cut out.
>> In 1970, Taylor founded the Computer Science Laboratory (CSL) of the Xerox Palo Alto Research Center (Xerox PARC). Through the 1970s, CSL researchers became known worldwide for a number of important innovations necessary to the creation of the Internet. CSL invented and built Ethernet, the laser printer, and the PUP (PARC Universal Packet) protocol. PUP was introduced seven years in advance of the implementation of the Internet protocol, TCP/IP. Within Xerox, all of these technologies enabled the construction of the first internet.
Categorisation of a data set like this can be non-trivial and will false positives. We can cherry-pick such a false positive and ignore all the actual positives. That said, thanks for this feedback - this is exactly what is needed to improve status quo. We are on it and will report back in a few hours with the category split I suggested in the previous paragraph. Do let me know if you think it can be broken down into a different type of granularity?