As expected, Webflow Conf 2024 delivered some exciting news on the features being released this year, with a few cutting-edge developments like the newly integrated Optimize feature.
However to the surprise of many attendees, very little of the new functionality is included in the base Webflow platform, even in a limited free-tier or trial-plan fashion.
Moreover, those prices are much higher that we're used to seeing in Webflow's self-serve sub-Enterprise plans, and that price must be paid before you can even examine the feature on your site.
It's a big jump, as well;
The subsequent sticker shock led to a number of frustrated posts following the conference on Reddit, Facebook, and in Webflow's forums, which can be summed up by this forum question;
How do I justify this to a client, especially when we’re talking about small businesses?
Here's my personal take on this.
Rethinking SaaS costs
Many freelancers make the mistake of evaluating costs in a personal-finances context, e.g. "would I pay / could I afford $X". I know this, because I did it for at least a decade, even when evaluating solutions on behalf of a client.
I saw it as part of my responsibility as a tech adviser to recommend services that I saw as low-cost or "affordable."
But successful businesses don't think that way. To a business, a $379/mo Optimize cost is a couple of sales, or 15 mins of a doctor's work, or a single premium car wash.
Especially once a business has employees, they're paying $50k to $100k per average employee. The website is an employee too... and it has a very specific job- sales and information.
If it does its job very well, It's exceptionally worth it.
If you're making that classic error in judgement, let's re-calibrate that thinking...
To a business, a cost is not a loss. A cost is an investment, and like any investment, it needs a return to justify it.
With this perspective, the question "how much does Optimize cost" is irrelevant until you also ask the question "how much revenue can Optimize generate?"
Before I tackle that question, I wanted to see how Webflow's Optimize pricing stacks up among its peers.
Webflow's Optimize v. Industry Pricing
Compared to the hosting fees you're used to paying, the sticker shock is real, and Optimize looks ridiculously overpriced... but when you look at Industry pricing, you'll form a very different picture.
Here are some examples.
Optimizely doesn't publish pricing on their site anymore. They've replaced it with a "request pricing" button. Google and other sites indicate-
For Optimizely, that $36k USD is the starting point.
There are many other choices, Webflow's blog has an article on this too. In general, the pricing is not cheaper, though some like VWO and Sumo have a free plan to trial the product, which I personally require when setting up a new client...
My job is to identify and evaluate solutions. My clients need to see what they're getting, and I need to know if it meets their requirements.
Who is the Customer?
Businesses that invest in A/B testing are generally already industry leaders, who want to maximize the performance of their existing site. They've done the basic work... design investment, product investment, messaging, SEO, CPC... and now they're at the stage of-
How can we squeeze this juicy orange just a little bit more?
This is known as Conversion Rate Optimization ( CRO ) and it's where A/B testing comes into play. You're now at the point where improvements mean swimming in data, and running "experiments" to determine what can work better.
At this level, business are generally successful, and willing to pay for that added advantage.
What if my Clients Can't Afford It?
When clients balk at the price, your best bet is to look for other platforms that offer the feature / price balance they're looking for. There are a lot of them.
If you know how to code, the options grow dramatically because you can build your own solutions based on Zaraz or Posthog which are cheap but developer-oriented.
I want to add an important note here-
Nocode is expensive.
It's expensive because the code doesn't disappear, it's just engineered into the SaaS platform rather than built by your dev team and supported by your ops team. It's a convenience, and often a pricey one.
If you want low cost, learn the skills to code and engineer your own solutions.
The trade-off is, you pay those feature-delivery costs in hours rather than dollars.
Why I'm Excited About Optimize
If you've done much A/B testing, you'll know why this is exciting.
Optimize promises-
- Full designer integration for the variant setup, which means easy, clean setup
- Efficient, server-side variant delivery
- Parallel multivariate testing. Demonstrated in the keynote, the possibility of efficiently running multiple experiments at the same time is very appealing
- Dynamic ongoing optimization based on user behavior, to keep your site at the top of its game
And I'm also excited about what the strategic direction of the "Experience Platform" means for Webflow. Optimize has the potential to turn Webflow into something incredible.
Imagine if...
- Your one website could auto-optimize for multiple markets, multiple languages, and multiple locations, to deliver the best message for each culture
- You could auto-optimize for device types and breakpoints, to deliver the best experience for each device
- You could auto-optimize by demographics like age group and gender. If User Accounts and eventually ECommerce could support FB oauth login and Google oauth login, there's some real potential for an epic experience for customers.
- Your experiments were AI-generated, based on what Optimize had learned about your user behavior. A tweak in color. The resizing or repositioning of an element. The timing of a popup display. As your user behavior data grows, so can your user intelligence, and your UX decision-making. Today I'd want to have Optimize intelligently recommend experiments for approval. Tomorrow AI might be solid enough to just let it make its own decisions in realtime.
5 years from now, maybe we'll just hand a site some brochure scans and a dropbox of images and video... and the entire site will be automatically generated and optimized.
But not today.
The Key Problems
Today, we're here are the central problems we're seeing with Webflow's initial release of Optimize. Hopefully these won't be here for long;
- No trial, which means that clients can't try the feature on their site
- No free-tier, which means that freelancers & agencies can't explore and learn the capabilities of these new features, and therefore can't sell them
- Annual billing. Add-ons are tightly bound to your site plan which means that if you are on an annual site plan, you are currently forced to purchase the add-ons annually as well. That's a big gamble on something you can't test.
- ECommerce is not yet supported by Optimize. Even if you've bought a template that had ECom enabled, and you're not using it, you're likely excluded. Moreover, ECom clients are likely to be the group who would benefit most from Optimize, since it can directly translate to on-site sales.
Feature ambiguity
Things we can't know until we use the product...
- What exactly do Optimize's reports include?
- How long do experiments run? When Mary visits, how long is she connected to her experiment group?
- Are experiments site-wide, or page-specific? If one experiment is to put a "Buy now" button in my nav, will that same experiment be consistent site-wide?
- How are users tracked? Are cookies required, or is a server-side fingerprinting mechanism used? How accurate is it at unique identification?
There is a corollary question here as well, which is "What is the current state of GDPR regulations regarding cookieless tracking?" While it's not specifically a Webflow question, it's an important part of the designer education surrounding any A/B testing product.
Pricing ambiguity
This part, I'm still trying to wrap my head around.
Most A/B platforms I've explored use a billing metric of "monthly tracked users" or MTUs. In most cases MTUs represents a "budget" that you want to impose on your A/B experiments, i.e. how many people per month you want to experiment with.
I'm not seeing MTU's mentioned in Webflow's docs, which only mentions page views - and that adds some further confusion around pricing.
I'll present these as open questions, which I'll hopefully be able to answer soon;
- If a client pays for 10,000, are they getting 10k experiment results for that investment, or only a portion of that?
- Can a client restrict an experiment's reach and pay for a subset of the traffic they want? Page views suggests that the cost of Optimize is directly tied to your total site traffic, and that you may be charged for 100% of users visiting.
- In this case, what happens for clients that have severe traffic fluctuations? Do they suddenly get a billing increase?
How this hurts Webflow
All of these points are challenging for us as an Agency but I think it's just as important to consider the impact of these gaps on Webflow's success.
Freelancers and Agency partners can't sell what they can't see
This is where I feel Webflow sales could do a much better job of equipping and training freelancers and agencies in the Webflow platform. The big problem is that we can't sell what we can't see. Currently there's no way to learn Optimize, test it, see results, give a client a tour of an Optimize-enabled site...
Annual plans "break" the accessibility of add-ons
Arguably, Webflow's best hosting customers are on Annual plans and Webflow benefits from the "lock-in" of that level of commitment.
However the way add-ons are priced, those "best" customers are punished when they want to trial a new add-on feature like Optimize, because they're forced to play a large up-front bill before they can use it.
Customers are beginning to distrust the Webflow platform because of the lack of billing controls
The recent change to bandwidth limits is a good example.
Some small clients would rather their site be shut down temporarily if it hits its bandwidth limits, than get a much more massive bill- but there are no safeguards.
Optimize for All? Someday, maybe...
My personal preference would be that Optimize was available and accessible to all Webflow hosting plans... and that it were one of the central value propositions of choosing Webflow as a platform.
It has that potential.
Imagine a world in which every website built on Webflow were able to use and learn Optimize fully- but that they were limited to one experiment at a time, and maybe only certain types of experiments like a simple text change.
It's simple, but I think this would go a long way towards selling a great feature.
