Articles, WordPress

WordPress Plugins: SAAS! Gota Go SAAS. Right Brah?

All the things meme. Text is "SAAS all the things! All software is a service! $7.99/month! For the last year or so, I can’t get into a discussion about the WordPress plugin business without hearing why I’ve got to go SAAS, seriously SAAS, get that recurring revenue. I’m constantly hearing about why I need to turn my plugins into a software as a service (SAAS) model or create a SAAS business.

This worries me. The SAAS obsession can leads to businesses designed around a business model, instead of fitting a business model to a solution for a specific pain point. Also, the got to go SAAS brah push is paved with promises that it might not be able to deliver, depending on the product.

Don’t get me wrong, there is a lot of upside to a SAAS for developer, business and end user. So I’m not anti-SAAS or incorporating some of its benefits into a traditional plugin. I just think we need to break down why these things do or do not make sense.

Get Monthly Recurring Revenue

They want you to think you can’t charge for your plugin monthly

A lot of time I hear that I need to make my plugins a SAAS business so I can get monthly recurring revenue. This argument has at least three logical flaws.

  • It assumes you can’t charge for a plugin monthly

You can. For Ingot, we offer monthly and yearly recurring payment options. We use Freemius for this, it’s pretty painless for us and the end-user.

  • It assumes monthly payment options convert better than yearly payments

Ok, this is generally true, but if it’s not you’re leaving money on the table. If people are just as likely to pay you $100 a year, as $10 a month, you’re probably better taking that $100 up front, even if you get $20 less a year. The hope that I will have $10 next month doesn’t pay my bills. $100 now does.

A ton of MRR is awesome. But a little bit of MRR for a growing business, in terms of running that business, isn’t as cool as actual monthly revenue.

  • It assumes your plugin has ongoing value

Some plugins are more of a use once or twice type than essential type of plugin. These plugins — importers and converters — should always use one time payment.

Vova from Freemius has explored all of these points in alot more detail on their blog.

The big thing to consider is whether we normally make plugin licenses just about support and updates, not ongoing service, because that’s how we normally do it or if there is an actual reason for it. Consumers are used to paying an ongoing fee for services. I pay for Netflix every month. If I didn’t pay, I wouldn’t get Netflix.

Ingot doesn’t break sites when the license isn’t paid. It just always chooses the first possible variant and doesn’t record any data from the test. That way Ingot still creates valid HTML on the page, your site isn’t broken, but you don’t get the benefit of our algorithm.

That’s a lot nicer than Netflix. Last time I forgot to update my credit card details in Netflix, I logged on and they had removed all of the features besides the billing screen from the app. I didn’t get a single episode of House of Cards! Not one!

Yes, the licensing system in Ingot is super easy to defeat. But, people who don’t pay for software are not paying for my software whether they use it or not.

So, whatever, I’ve got a lot of things to worry about in life. Seriously, Donald Trump might be the next president and you want me to worry about people not paying for software. Meh…

Going SAAS purely because you can charge for your product in a specific way, not because you can deliver a better product via SAAS, or it fits better to the problem you’re solving or your solution, is bad business planning.

Protect Your IP

They want you to think your IP needs protecting

Photo by: Leigh KendellThis is the weakest argument for SAAS all the things in my opinion. Seriously, give me any good idea for a WordPress plugin and a small amount of time and money and I will reverse engineer it. So can a thousand other developers.

Yes, everyone can fork my plugins and use it anyway they want. I just don’t care. People try and get me to care, and I just don’t. When we launched Caldera Forms we didn’t fork Gravity Forms, we made our own thing and made it different, that is our advantage, it’s different.

I strongly believe that the ability to openly discuss source code with other developers leads to better code. Free software is both an ideological principle and a pragmatic way to make better software.

Does this apply in every case? Probably not. Getting investors interested in backing commercial free software isn’t easy. It’s hard for them to see what they should pay to make free software.

If you work in a free software space and you can’t convince yourself or an investor of the value of open source, one or both of you shouldn’t be involved in WordPress.

Serve One Stack

They want you to think some other plugin’s error is your problem

Photo by: Paul JarvisAs a developer, not having to worry about every single version of PHP and random crappy shared hosting configuration is a one of the most attractive propositions you can offer me. I dream of having one server stack to debug and being able to log every error. Being able to reproduce and test errors on the one single stack my app will be deployed on, that’s a wonderful thought.

iFraming in a plugin’s UI and not letting all the random crap that happens on a WordPress site break your app should make for less support.

I am making a lot of assumptions in this section.  Everyone of them needs to be evaluated to see if it is really true, not hopefully true.

But even if they are, that is not enough of a reason to go SAAS. Sure, running the processing on your server versus the user’s server might be the only way to make the service work. This is a case by case evaluation to make, but I’m just saying that Moore’s Law is against you on this.

But here is the other big issue: operational cost. Ingot runs totally on the end-user’s server. If someone uses it to run a million tests a day, or five tests a day the cost to us is $0. If we were a SAAS that cost might be very different.

This gives us serious flexibility in designing our pricing model. Assuming a fixed support cost, we could stop charging for the plugin tomorrow if we felt upselling to services was a better model. Our costs of running the core service wouldn’t change. Not being a SAAS product gives us major flexibility, including being a SAAS product at some point in the future.

For now we provide Ingot for between $0 and $100 a year and the risk of having the wrong pricing model doesn’t include running up a ton of AWS usage we can’t afford.

One of the big advantages of the WordPress plugin industry vs being a SAAS is we don’t pay to run the servers are software run on. Are you sure you want to give up that advantage?

Data Control

They want you to think “regular users” don’t care about owning their own data

Photo by: Aleksandra BoguslawskaWordPress is about democratizing publishing. As WordPress evolves that becomes more abstract, but the fundamental idea is that you control your data and what to do with it or not do with it. There is value in this and it is one of the reasons that people choose WordPress.

This point is not just ideological, it is pragmatic. WordPress users expect hooks, to customize how plugins and themes work. If the plugin doesn’t run on their servers, there are no hooks. Targeting an element in a cross-origin iFrame via CSS or JavaScript is not possible.

The more restrictive your APIware plugin is, IE the more you protect your IP, the more you have betrayed the WordPress promise of ultimate customizability. You start to invalidate the choice of WordPress for a project. And yes, WordPress isn’t right for everyone, but still, I want it to be as much as possible because WordPress is how I make a living and how so many people I know do.

The more complicated a site is, which often corresponds roughly to budget for the site, the more likely a site developer needs to modify how a plugin works or query its data directly. The more you close this off, the less useful your tool is to those who make WordPress sites professionally.

Professional developers and site builders are very important to your business. They recommend products, they blog about products, they buy multi-site licenses. They make other plugins that might integrate with your plugin.

SAAS Makes A Lot Of Sense

They want you to think Josh is anti-SAAS

Crop of ant in flower

Let’s talk about Akismet for a second. Akismet is Automattic’s anti-spam system. Without it would be covered in spam and would have failed. Without a healthy selection of anti-spam solutions all WordPress sites would be covered in spam and spam would have failed.

Akismet is one of Automattic’s oldest products for self-hosted WordPress sites and it uses a freemium SAAS model. The WordPress plugin is APIware that passes a suspect comment or other content to Automattic’s servers for a spam or not spam determination. The actual anti-spam system isn’t in the code they deliver to you in the plugin.

I don’t know for sure, but I suspect that if the that algorithm was running on your site it would be very anti-performant. I also suspect that in order to keep on top of increasingly intelligent spam bots, they need to crowdsource their machine learning.

Both of these factors probably make for an infinitely better service. In the process they got around having to release their code under a free software license and created a stream of recurring revenue. Both of those things I assume please the venture capital firms that Automattic would not exist without investment from.

And I pay that $5 a month for many of my sites. Great service, totally worth it.

So, will you see a SAAS service from me? Maybe. There are some interesting ideas being prototyped that might overcome every objection I’ve given here for going SAAS.

My team is way better equipped to deliver WordPress plugins than SAAS services — no devops here — so it makes me a bit nervous. I would suspect you would first see us offer add-on services to our core products that make them better or easier in some way.

People who skim this article are probably going to think I’m anti-SAAS for WordPress products. I’m not, I’m just saying that this business model has to be chosen because it’s the right business model. When the conversation starts at business model, before even knowing what the business is, that makes me antsy.


  1. Josh, those are all very valid points. I’m looking into a SASS project myself, but it does seem that every podcast, and development article screams SASS- quality and security are the main things that should be considered first- I’m not convinced that those 2 are tantamount to alot of the new SASS applications popping up.

  2. Great article Josh. It’s easy to go SaaS by default, cos all the kids are doin’ it, and because (as you say) it makes a lot of sense for many plugin businesses. But you’re right, it’s not always the best option. And yeah, talking business model before understanding the business is a worry.

  3. Thanks for writing this. Event espresso, the plugin I work on had a saas site called event smart.
    Saas users are often less technical, but also seem to be more temporary customers. They’ll usually use our saas site to sell tickets to just one yearly event, whereas self-hosted users usually use our plugin to sell tickets to dozens of events. I think they’re slightly different customer segments, and just saying “saas all the things” is quite a narrow view, in my opinion.
    I personally avoid signing up for saas services myself, precisely because of the recurring expense. It also stinks being at the saas company’s mercies: will they go out of business? Change owners? Change from charging a flat fee to a pay-per-word model (anybody remember Movable Type? Wasn’t that a saas model business, that changed their pricing scheme, angered their customers, and had them flock to WordPress precisely because it WASN’T saas? Despite even though it was probably an inferior product at the time).
    I agree saas has a place, but it’s not without difficulties of its own.

  4. “People who skim this article are probably going to think I’m anti-SAAS for WordPress products. I’m not, I’m just saying that this business model has to be chosen because it’s the right business model. ”
    I think you should put the paragraph above into it’s own section such as conclusion or somehing. As you said, people reading your article may think you are against SaaS and most probably won’t read the whole thing.
    We offer a PaaS product as well as regular WordPress plugins. To me, one of the more important reasons you want to go SaaS vs. self-hosted is the complexity of the algorithm you offer. If it is somehing that can not be implemented in the client machine, I would go for SaaS or PaaS. The other arguments, in my opinion, are all business model related issues which can be resolved one way or other.

  5. Augustin

    This a really interesting article, deep diving into SaaS roots and consistency vs. services offered. As the co-founder of a freemium SaaS plugin for WP, I would add that if your service offers additional value to users vs. historical solutions, plus, you deliver it on a API basis, then it makes sense to me to adopt a SaaS model.
    At the end, as you mentioned it, the question is: does the service delivered through SaaS is better than non-SaaS solutions?
    Thanks for this article and happy to discuss further.

  6. I love this article and your approach towards this issue. Everyone should make a wise assessment of their product’s options and base the decision whether to go SaaS or not on that. Going SaaS should not necessarily be the default option for every product out there.

Leave a Reply

Theme by Anders Norén