Nick Samuel

Sloth Cloudflare Worker Generator – Review

Edge SEO Served with Extra Salt

The concept of Edge SEO sort of blew my mind slightly when I first read about it.

It resonated with my heavily after I had just gone through a challenging SEO period where we had landed a new client but their website was for want of a better phrase… absolutely shite.

Thankfully they knew it themselves, but this didn’t change the fact that we would still be expected to make SEO inroads in the interim (6 months) whilst their new website was being developed and put live.

Naturally the relationship with the incumbent dev agency had somewhat deteriorated.

However we the good SEO guys still anticipated that if we made reasonable SEO requests, writing in polite and reasonable tones , set to realistic and reasonable time frames, then we’d have a reasonable chance at making reasonable SEO progress.

Obviously this didn’t happen and instead we were stuck treading water. And then the website was delayed. Again.

Would having access to something such as saved us here?

Let’s just say it certainly wouldn’t have hurt…



For those out of the loop, Edge SEO is a phrase attributed to technical SEO wizard Dan Taylor, and the agency he works for, called Salt.Agency.

It gained popularity towards the tail-end of 2018 when Dan and his partner, Igor Krestov (Lead Developer at Salt), won a research award at Tech Boost – a Technical SEO conference hosted by Catalyst Digital In Boston, USA.

The title of the award winning paper? Utilising Cloudflare Workers to Overcome the Challenges of Legacy Tech Stacks and High DevOps Costs.

Sloth is the name of the tool which Salt built to generate worker ready Javascript code for your website based on a series of WYSIWG fields, toggles and check boxes.

I believe it has been open for Beta testing at least since Tech SEO Boost, but now the guys have opened it up to public sign-ups.

So without further ado, here’s my “review” of Sloth.Cloud.

And by review, I of course mean I play with it for a few hours and give you my initial thoughts.

Current Features

At the moment there are six features available in the Sloth App:

  1. Robots.Txt
  2. Redirect Manager
  3. AB Testing
  4. Hreflang Manager
  5. Security Headers
  6. Logging (generating server logs on the Edge)

Potential Features:

If the current version was the only version of the tool to exist then I would be ecstatic, however Dan and Igor have already begun to tease us with future features.

From reading the Sloth documentation here and browsing the main site, it looks like the following features *could* be in the works:

All I can say is damn, if all these options ever get released then Sloth might just… Edge…out Keywords Everywhere as every SEO hipsters favourite free SEO tool. Including mine.

What I like

I hate people who literally reference Pareto’s Principle and religiously repeat some bollocks about 20:80.

However, I do like the general concept behind what I think Signore Pareto was getting at…sometimes less stuff matters than what you initially think.

And in the case of Sloth.Cloud all I need is the robots.txt editor and the ability to manage redirects on the Cloud and I am a very happy bunny.

That’s 2/6 or 33% of the current functionality.

If Sloth just did these two and allowed you to deploy to CloudFlare, it would be regarded as a substantial innovation.

The fact is though, it does so much more.

And as I write this even more features are currently being tested in in beta. This means Sloth’s final form will be ridiculously OP’d.

Most of all I love the idea that Sloth gives us Technical SEOs another powerful tool to our armoury, another option to consider, another way of problem solving.

Genius doesn’t have to be complicated and innovation doesn’t always have to mean a new idea.

Salt didn’t invent CloudFlare workers but they identified the opportunity for SEO and introduced it to a whole industry with a free to use tool.

They deserve a tremendous amount of kudos for this and I defy any SEO technical or otherwise to read about Edge SEO and not get at least one lightbulb moment.

What I don’t like

Ok, so what don’t I like about Sloth.Cloud? Nothing honestly, but nevertheless I will outline a few of the problems I can into when I was playing around with it.

Just to clarify, yes my “tests” are completely subjective and might not be repeatable. I also “tested” this for maybe half a day max and <10 deployments.

And yes, by “test” I mean I played around with it; I’m not claiming there’s much scientific rigour here whatsoever…

Route/URL Ambiguity

First of all let me start by saying that the generation of Sloth code is great. Deploying with the API key is an absolute breeze and worked brilliantly for me every time.

However, I don’t think that Sloth gives you the option to only deploy certain workers on certain pages.

Or more accurately if you do go down the manual copy and paste route which I did a few times and then delete the default routes (or modify them), then Sloth will post the worker script but it won’t actually be active.

A Cloudflare route is basically the rule which will pattern match the worker scripts to and you may be familiar with it if you’ve previously set up any page rules using CloudFlare.

Perhaps there could be an opportunity to add one or two bits of documentation on creating routes, Sloth side, for the avoidance of confusion and to save us a small amount of Googling.

Potential for conflict

Not a criticism of Sloth.Cloud at all, but just another reminder of the dangers that when you live life on the edge, you need to be careful not to cut yourself on the edge, because the edge can be pretty sharp.

Before I gave Sloth a go I was playing around with a Cloudflare app called Logflare which also uses service workers. It collects logs and tails them to a pretty cool looking interface, but ultimately isn’t a long-term server log collecting utility.

Note: I haven’t had time to test Sloth’s log collection yet but it looks like it has potential.

I first thought I must have deployed my Sloth app script incorrectly as I chose to implement them manually at first but even with API deployment I had Javascript errors.

My page refused to stop loading and it had a loading icon whirling in the browser tab with GT Metrix and other site speed tools unable to time the fully loaded time (as it never stopped loading).

Anyway,  I removed the app and the Sloth code ran absolutely fine as you would expect </cautionary tale>.

SiteSpeed (Theory)

One of the biggest theoretical issues with Edge SEO is that it could or should have a negative, or at a very minimum, a very slight impact on site speed and performance.

Rendering JavaScript before serving the modified page back to the browser (which then might also have to render JavaScript) is of course an extra step.

As per this post on my homepage sitespeed, this page is approximately 1mb with 10-11 requests and a fully loaded time of 363 seconds.

And during my tests I’m pleased to confirm that Sloth had pretty much zero negligible impact on this. Damn, I shouldn’t have wrote this subheading in advance…

I don’t want Sloth or Edge SEO to be slow just to prove a point, but I still need to caution that my Sloth script was pretty light and didn’t exactly features 1000s of redirects etc.

Ultimately, I suspect in most cases if you need to resort to using Edge SEO in the first place, then you will more than happy take the extra 200ms or so extra load time.


Sloth is a brilliant concept with a fantastic user friendly execution, and something that I’d advise any (Technical) SEOers to play around with.

Edge SEO, Sloth and CloudFlare won’t revolutionise how the average agency does SEO on a day to day basis, however it does provide much needed tools in the overall SEO arsenal.

Sloth directly tackles an age old (agency) problem; sometimes accessing and making changes to client’s servers isn’t possible.

Depending on how Dan and the team develop the next version of Sloth, it could very well have a similar powerful but low-key impact as Google Tag Manager.

Overall it’s not very often that we see genuine innovation in the SEO sphere, so I think it’s important we recognise it when see it.

Nice one, Salt.