A few months ago, I came across this article by a SaaS founder detailing how he increased his product’s signup-to-paid conversion rate by 700%. It immediately caught my attention because the jump seemed too good to be true.
Here’s a TL;DR version:
-
- Jon was looking for ideas to improve his product’s low conversion rate.
- After getting mixed feedback about the onboarding flow, he realized that the real culprit might be the signup quality.
- He tested this hypothesis by adding friction during the signup flow.
- The number of signups dropped significantly without affecting the number of new customers, resulting in a 8X conversion rate.
Like I suspected, the article was a bit clickbaity because the total number of conversions didn’t actually increase. But it did highlight one important point: Understanding whether you have a conversion problem or an acquisition problem by examining the quality of your signups.
Too many SaaS companies get tricked by the overall conversion rate without realizing how much it’s influenced by signup quality. They waste time optimizing their onboarding for users who are poor fits while degrading the experience for their ideal users. This story perfectly illustrates why you shouldn’t do that.
However, the article did come with a few risky assumptions. For example:
-
- It didn’t look at changes in top-funnel traffic, which is important for comparing how much the signups dropped.
- Based on the trial limit, I estimate that the conversion window needs to be set at a minimum of 7 days. The conversion rate calculation used in the article doesn’t seem to be time-bound, which makes the before-after comparison unfair.
But the point I want to discuss today is a conclusion that I know many will incorrectly make from the story: “You should eliminate all bad signups.”
While it can be a good move in some cases, it’s often an overcorrection for high-volume, mass-market SaaS products taking a product-led growth approach. Allow me to explain why.
Definition of bad signups
Let’s first define what we mean by “bad” signups:
These are signups that fail to meet your minimum qualification criteria, based on firmographic, demographic, use case, source, and intent data collected via onboarding survey, enrichment tools, sales calls, pre-signup activities, and product usage.
They can still become paying customers and even make up meaningful MRR, but data shows they significantly underperform qualified (aka “good”) signups across key metrics — making them poor targets for deliberate acquisition.
Where you draw that line depends entirely on your context. You can also opt for a lead scoring approach vs. a binary qualification.
The goal of eliminating bad signups
The practice of actively weeding out bad leads originated from the sales-led era of SaaS (and well before SaaS).
Because a sales team has limited bandwidth, they must prioritize leads with the biggest deal size and highest win probability. Without a smart self-serve flow to convert signups into paying customers on their own, every bad lead becomes a major distraction.
Back then, new leads came mostly from paid lists and targeted outbound campaigns, resulting in a high cost per lead. The more bad leads you had, the more marketing budget you wasted, and the less efficient your sales team became. That’s why a high proportion of bad leads was synonymous with slow growth.
But if you look closely, some of these factors don’t apply to PLG companies.
How are product-led companies different?
Let me share a story to illustrate why PLG companies should not rush to eliminate bad signups.
A SaaS startup I advised (>$2M in ARR, SMB-focused) was struggling to keep up their growth momentum. Despite having a growing, mostly organic top-funnel, its signup-to-paid % was declining and churn was increasing. I ran a diagnosis and found that:
- Their ICP actually contained two distinct user segments. After breaking down the data, it became clear that one of them wasn’t as good of a fit as previously thought.
- The number of bad signups were growing faster than the good ones, dragging down the overall metrics.
- The sales-assisted flow converted 4X better than the self-serve flow, but only a small portion of users had entered it.
Based on these insights, the team set two goals: Increase the overall quality of signups and the number of sales-assisted leads. I shared some ideas about how to better acquire and convert their newly refined ICP and left the team to execute.
When I checked in after a quarter, the company’s good signup ratio increased significantly, the number of sales-assisted leads per month grew 8X, and the signup-to-paid % also increased by 250%. Sounds like a big win, right?
But as the founder excitedly shared the progress, I spotted something odd: their new MRR was actually shrinking.
As it turns out, what they did was replace the sign-up button with a demo request form. This tanked the visitor-to-signup %. Essentially, only high-intent visitors would bother to book a call, and only those who qualified would be given a trial, so of course the signup-to-paid % increased. While this change probably helped convert some signups who would’ve dropped off without any sales assistance, it didn’t make up for the loss of self-serve MRR. All things considered, the experiment was a net failure.
This is a common fallacy among SaaS companies. Yes, your secondary metrics (activation, conversion, retention, ARPU, etc) will look better if you find a way to filter out low-quality signups from the start. But unless it helps you acquire better users or capture more value per good signup, you’re simply sugarcoating the data.
Let’s look at another example. What if I told you that a company improved its signup-to-paid % from the left to the right.
Impressive, right?
But what if I told you that they did it by killing the free plan, removing the “Powered by” watermark, shutting down the affiliate program, and deleting SEO pages targeting low-converting (but high volume) keywords — all cost very little to maintain — and this is what happened to their new MRR?
Not so impressive anymore, is it?
This is the main difference between a product-led and a sales-led growth motion. When you have self-serve mechanisms and low-maintenance acquisition channels that don’t cannibalize each other, eliminating low-quality signups isn’t always necessary.
In fact, doing so could hurt you in many ways.
How can it hurt?
Wasting effort that could be spent on acquiring better signups
You can improve your signup quality by either:
1) growing good signups or 2) reducing bad signups. Both paths can lead to the same quality ratio, but growing good signups is what drives real growth.
Every minute your team spends preventing bad signups is a minute not spent acquiring good signups. This compounds into a major opportunity cost. Although reducing bad signups has its benefits (more on that later), it should rarely take priority.
Filtering out good signups by accident
When you add friction or cut acquisition channels to eliminate bad signups, you’ll inadvertently lose some good signups in the process.
For example, requiring a company email at signup might drive away high-quality, mid-intent prospects who want to explore your product privately before triggering an official trial.
This is perfectly fine if it helps you unlock opportunities to convert more users overall. But if that isn’t the case, then you’re just leaving revenue on the table for no good reason.
Missing out on extra cash flow to fund your growth
Not all MRR is created equal, true.
But still, revenue is revenue. As long as you know which portion of your MRR is sustainable and focus on growing that, you don’t have to say no to bad MRR.
Instead, you should use that extra cash flow to fuel your good MRR growth by investing in activities that can be dialed up and down as needed, such as paid ads, influencer sponsorship, and freelance content production.
Killing your ecosystem before it forms
Companies like HubSpot, Shopify, and Notion have proven that building an ecosystem around your product can pay off big time. Some of your users will never be monetizable, but they can still bring incredible word-of-mouth, network effects, and brand awareness.
These benefits take time to compound and are not easy to attribute. If you only care about users who can be monetized directly, you likely won’t have the critical mass to form a meaningful ecosystem.
Missing opportunities to learn
Your definition of an ideal customer won’t stay the same forever. It will evolve with your product, market, and learnings. By pre-filtering every signup that doesn’t meet your current criteria, you create a self-fulfilling prophecy that prevents new insights.
“But wait, didn’t you say that we shouldn’t lump all signups together when analyzing data?”
Yes, that’s still true. But there’s a difference between segmenting your data and not tracking an entire segment altogether. It’d be wise to leave some room to be proven wrong.
Enjoying this post so far? Subscribe for more.
When to actively eliminate bad signups?
Having said all that, there are still many valid reasons to intentionally minimize the inflow of bad signups.
Shortening CAC payback period
If you have variable costs that scale with each new signup, you better make sure every dollar spent gets returned as quickly as possible (measured as CAC payback period).
The variable costs I’m talking about aren’t limited to paid marketing — server, API, tooling, and personnel costs that go toward supporting your freemium and trial users should also be included. Traditionally, these are categorized as COGS (cost of goods sold), but they’re effectively acquisition costs if you think about it.
The more bad signups you attract, the longer your CAC payback period gets, and the more strain there is on your cash flow. In this case, reducing the number of bad signups does free up your resources to help you grow faster.
But then again, whether this should be a prioritized effort depends entirely on your company stage, runway, current CAC payback period, and GTM strategy. For most early-stage startups, you should focus on finding scalable ways to grow your target users. Doing so requires experimentation. Many experiments will fail, some will win. You won’t know what “good” looks like without some comparisons. This is why CAC optimization makes more sense after you have a large enough sample size.
Blocking fake signups
There are bad signups that convert poorly, and there are fake signups that exist to exploit your business — bots, spammers, trial abusers, fake referrals, etc.
These fake signups don’t just create operational headaches; they pollute your data and skew your decisions.
In my experience, at least 10% of signups at every post-PMF SaaS company fall into this category. Case in point: Clay recently did an analysis and discovered that 47% of their signups were fake, forcing them to implement a proper detection system.
While fancy tools aren’t always necessary, you should still have basic protections like email verification, CAPTCHA, and usage monitoring. Better yet, design your feature gating experience and affiliate program in a way that discourages fake signups in the first place.
Providing better support for your target users
It’s a universal law in SaaS that your worst users will demand the most support. This steals precious time your support, success, and sales teams could otherwise spend on converting and retaining your target users.
Even though you could prioritize which users to help first, getting that prioritization right is easier said than done. If a large number of users don’t receive enough support, word will get out to hurt your future growth.
The reality is that SaaS users today have high expectations, even on a free plan. If you can’t meet those expectations, you’re better off limiting who you let into your product in the first place.
Changing your brand perception
Your brand is shaped by who your users are and how they talk about your product. When your target audience sees many users they can’t relate to, they begin to question if your product is truly built for them.
If this sentiment comes up repeatedly, it’s time to weed out users who might misrepresent your brand, even if it sacrifices some short-term growth.
More often than not, this hints at some deeper issues with your positioning. It might be too vague or fundamentally flawed. Fixing that root problem should be your top priority.
To sum it up
For SaaS companies taking the PLG route, my advice is to view bad signups as a byproduct of your acquisition effort. If your product solves a big enough problem, chances are it’ll attract many poor fits and looky-loos, and sometimes they grow faster than your good signups.
Eliminating these bad signups should not be a goal in and of itself — growing your good signups and MRR is.
If getting rid of bad signups helps your company grow more efficiently, by all means do it. Just make sure you aren’t padding your metrics at the cost of all the indirect, long-term benefits you could’ve gained.
Having a mixed caliber of users definitely makes your data muddy. It takes skill to cut through the noise and maturity to accept that some of your metrics will always be inflated.
But that’s precisely what separates winning companies from the rest.
