Why self-service freemium capped LTV at Crystal, and how its PLG funnel now feeds B2B sales at about 4x LTV
Drew D’Agostino · Crystal
Published September 1, 2022Drew D’Agostino, Founder and CEO of Crystalknows shares the inside details on:
- Why our self-service model couldn't scale
- How we validated the B2B sales opportunity
- How we use our PLG funnel to drive B2B sales
Explore the lesson. Sign in to access the slide download.
Drew D’Agostino describes Crystal’s move from a self-service freemium model toward a higher-value business motion. He contrasts funnel growth with LTV and retention, then covers the operational trade-offs of serving larger customers.
Read the source passagesFind the ideas you need and go straight to the source.
“the new model and how that has played out so far over the last two years at the high level we've about 4X star LTV which is the main kind of metric that I've been using as success to make these decisions and build the team and everything so three stages so I'll start with just basic overview of the product um does anybody here familiar with Crystal Eric okay a few people so it's a um we call an Adaptive selling tool we serve mostly B2B sales teams and they”
“under her this is the key stat that I am laser focused on with our team it's improving improving our LTV so all throughout this self-service journey of our company I could not for the life of me figure out how to increase ACV significantly I could not figure out how to get the churn down so that resulted in a long long slog of struggling with Lowell TV we had a very exciting product on the surface but just the nature of you know this credit card”
“um how we did that this was a part of the organic funnel and how it was generating those thousand signups a day so first one was just multiplayer features I could skip over these details here but the second was SEO so that was a major part of growing that funnel most of our SEO is just these personality tests and all these types of opportunities that we found online so bottom line here is that growing that”
“unexpected things where it's if you're one of those companies that's going from self-service towards Enterprise or B2B this is one of the things that just surprised me I was expecting roughly similar usage and retention patterns but what we saw was a jump from like 55 retention rate to about 120. and that was pretty much the same product except when it was an individual use case just you lose half of them over a year versus”
“the main ones are around sales cycle and complexity so if you've ever tried that up Market motion if you're going the opposite direction it's it's kind of a nice it's a nice reprieve from those things but going out Market we definitely had to build out our just the overall apparatus in the uh to the infrastructure around compliance and legal in all of this so that's been a big big trade-off and also much greater seasonality this was one that I didn't expect whereas our Revenue used to be very very steady now we see those huge drop-offs in the summer come back”
“Most SaaS Cannot Scale by PLG Alone.”
“Crystal’s Revenue Journey: PLG => B2B”
“In early 2019, we saw opportunities to capture organic traffic from personality-related search terms (i.e. “disc test”, “enneagram test”).”
“Assessments and insights for individuals”
“How we acquire 1K organic signups/day”
Video
Life After Freemium: How we killed our business model - Slide Deck
Drew D’Agostino is the founder of Crystal, a company that helps people communicate more effectively with each other by developing personality detection software and email assistance tools. He previously co-founded and served as CTO of Attend.com, a venture-backed event management software company based in Boston, leading the product,engineering, and support teams.
Drew D’Agostino · Crystal
Andre Cvijovic · Referrizer
Ben Gabler · Rocket.Net
Chris Cunningham · Clickup
Chris Marentis · Surefirelocal
Craig Fuller · FreightWaves
Each slide with a short explanation of what it shows. Open the slide text for its exact wording.
26 passages
We built a new business model Founder500, Sept 2022 Drew D’Agostino Founder, CEO of Crystalknows … and increased LTV by 400%
The speaker’s central lesson is that most SaaS companies cannot scale through product-led growth alone. The deck traces Crystal’s move from a self-service model to a combined PLG and B2B approach.
Most SaaS Cannot Scale by PLG Alone. My biggest lesson from the past 7 years:
The talk follows Crystal’s shift from PLG through a pivot to a combined PLG and B2B model. It covers organic acquisition, profitability, validation of sales opportunities, the transition’s risks, and account expansion.
Over the next 20 minutes: How we acquire 1K organic signups/day I’m going to show you PLG MODEL PIVOT PLG+B2B MODEL How we achieved profitability after VC Why our self-service model couldn't scale How we validated the B2B sales opportunity How we de-risked the transition to the new model Unexpected challenges of moving up-market How we increased LTV by 400% in 2 years How we use our PLG funnel to drive B2B sales How we expand accounts from 3- to 6-figures
Crystal aims to make sales teams more effective by adapting communication to each buyer’s needs and preferences. The product mockup pairs a buyer profile with suggested ways to tailor a sales pitch.
What Crystal Does We make sales teams more effective by adapting to the needs and preferences of every buyer.
Crystal was founded in 2015, received early backing from Salesforce, HubSpot, and others, and became profitable in 2019. The slide lists a 28-person team and says its last funding was in 2018.
Crystal’s History Founded in 2015. Backed early by Salesforce, HubSpot, and others (last funding in 2018). Profitable since 2019. 28 employees.
Crystal combines DISC-based communication insights with a browser extension and free behavioral assessments. It also offers higher-touch training and coaching.
Crystal’s Products DISC-based communication insights. Chrome Extension integrates with LinkedIn, Gmail, Salesforce, etc. Free behavioral assessments (DISC, Strengths, Values, etc). High-touch training and coaching.
The revenue chart shows B2B rising sharply as self-service revenue levels off and then declines. The timeline marks the addition of mid-market sales, customer success, and enterprise sales before self-service was removed.
Crystal’s Revenue Journey: PLG => B2B Started Mid-market sales team Started CS team Started enterprise sales team Launched with invite-only self-service plan Launched freemium self-service model Removed self-service
The LTV chart climbs most sharply during the B2B plus PLG phase. The slide compares $350 for early PLG, $480 for mature PLG, and $2,000 for B2B plus PLG.
Under the Hood: LTV up 400% Early PLG: $350 LTV Mature PLG: $480 LTV B2B + PLG: $2,000 LTV
This section examines Crystal’s pure product-led growth model from 2015 to 2020, before the shift toward B2B sales.
Pure PLG 2015-2020:
From 2015 to 2020, Crystal used freemium pricing and had no sales or customer success team and no marketing spend. The slide reports $420 yearly ARPU, with most customers paying monthly.
2015-2020: Pure PLG Freemium pricing model $420/yr ARPU 80% monthly, 20% annual No sales or CS team No marketing spend
Self-service helped Crystal build an inbound funnel, a product-led organization, and an efficient operating model. The slide highlights rapid product feedback, low acquisition cost, recurring cash flow, and more than 1K signups per day.
PLG Was GREAT for the Early Stages… Forced to build organic growth channels. Our self-service model gave us 3 great advantages: INBOUND FUNNEL PRODUCT/UX EFFICIENCY Rapid product feedback and iteration loop. Insanely low CaC. Consistent 1K+ signups per day. Tons of great logos. UX-driven mindset throughout the organization. Steady recurring cash flow. Tiny team (<10) supporting thousands of customers. Brand awareness. Low-friction buyer experience.
Multiplayer features make free tools more valuable when coworkers join, creating an organic acquisition loop. Shareable outputs also give users material to distribute beyond the product.
Organic Funnel #1: Multiplayer Features Robust free features. Team-focused UX. Lots of value from inviting coworkers to the platform. Shareables: downloadable content for social media, email, physical display, etc.
Crystal saw a search opportunity around personality tests in early 2019. The chart shows organic traffic rising after that period while paid traffic remains comparatively small.
Organic Funnel #2: SEO In early 2019, we saw opportunities to capture organic traffic from personality-related search terms (i.e. “disc test”, “enneagram test”).
The funnel narrows from 500K+ monthly unique visitors to 30-40K signups, then 3-4K trials and 150-200 customers. The slide pairs those stages with a personality-test landing page and assessment flow.
Organic Funnel #2: SEO The Self-Service Funnel: 500K+ uniques/mo. 30-40K signups/mo. 3-4K trials/mo. 150-200 customers/mo.
The self-service model struggled with churn, low average contract value, and limited expansion opportunities. The slide links monthly subscriptions and small accounts with price sensitivity, few customer-success interactions, and large companies remaining on single seats.
…But We STRUGGLED to Scale Revenue. Mostly monthly subscriptions. With 100% self-service, we struggled in 3 main areas: HIGH CHURN LOW ACV $$$ ON THE TABLE $420/year ARPU, spread across thousands of customers. Many HUGE companies using self-service. Mostly single seats and tiny teams. 5-8% total gross revenue churn. 2-3% involuntary churn monthly. Limited room for customer success. Limited opps to build customer relationships. No route for enterprise users to expand. Very sensitive to price increases.
This section covers Crystal’s shift to a combined product-led and business-to-business model, beginning in 2020.
PLG+B2B 2020-Now:
Team subscriptions retained customers better, while larger companies were not expanding from individual accounts into team plans organically. The slide says PLG economics depended on organic signups and that ongoing consumer-like churn was wearing down the team.
Why We Shifted from PLG to PLG+B2B: Team subscriptions had much better retention than individuals. PLG unit economics only worked with organic user signups. Large companies were organically converting with individual accounts but were not organically expanding into team accounts. The slow, steady grind of B2C-like churn was burning us out.
Crystal hired sales and customer-success leaders, added a real customer-success process, and bundled training with B2B subscriptions. It also limited self-service accounts and identified sales opportunities in the existing customer base.
3 Key Moves in 2020: Hired a VP of Sales who did it before ($1MM => $10MM ARR). Hired a VP of Customer Success and started a true CS process, bundling high-touch training/coaching with our B2B subscriptions. Placed new limits on self-service accounts, and identified B2B sales opportunities within our existing customer base.
Word of mouth, SEO, and marketing feed a free account that can qualify into SMB, mid-market, or enterprise sales. The diagram distinguishes the paths by annual contract value, company size, and use case.
PLG+B2B Sales Funnel Mid-Market SQL $3-24K ACV 20-5,000 employees All use cases Enterprise SQL $24K+ ACV 5,000+ employees Primary sales use case MQL <$3K ACV SMB & Individual All use cases WoM ~20K/mo SEO ~10K/mo Marketing ~2K/mo Free Account Assessments and insights for individuals Free trial of premium tools
Compared with PLG, PLG plus B2B brings longer sales cycles, more complex reviews, greater seasonality, and more people-intensive work. The slide contrasts those costs with PLG’s immediate, simpler sales process and steadier operating rhythm.
Risks & Trade-Offs PLG PLG+B2B Longer sales cycle (6+ weeks). Complex sales process involving legal, compliance, security, and privacy. Much greater seasonality. More people-intensive (hiring and management are more important) Emotional roller coaster: Higher highs and lower lows. Immediate sales cycle. Simple sales process. Muted seasonality. Less people-intensive. Emotionally stable: Predictable, smooth curves.
The ARPU chart trends upward over time, especially from 2020 onward. The slide’s headline says ARPU increased by 300%.
ARPU: Up 300%
The revenue-churn chart trends downward from 2020 through 2022, despite several temporary increases. The headline reports a 71% decrease.
Gross Churn: Down 71%
The revenue chart shows self-service leading early, while mid-market and enterprise revenue grow later. By the end of the period, mid-market is the largest of the three segments shown.
Current Revenue Breakdown
A model that reaches $1MM may not reach $10MM, and moving up-market takes sustained investment and conviction. The speaker also argues that accepting a temporary decline can unlock growth and raise the company’s ceiling.
3 Key Takeaways: The business model that got you to $1MM may not get you to $10MM. Going up-market requires long-term investment and conviction. Overcoming the fear of “number go down” can unblock your revenue and raise the ceiling of your company.
The recap returns to the journey from a PLG model through a pivot to PLG plus B2B. It summarizes the organic funnel, the case for changing models, and the lessons of moving up-market.
Over the last 20 minutes: How we acquire 1K organic signups/day I showed you: PLG MODEL PIVOT PLG+B2B MODEL How we achieved profitability after VC Why our self-service model couldn't scale How we validated the B2B sales opportunity How we de-risked the transition to the new model Unexpected challenges of moving up-market How we increased LTV by 400% in 2 years How we use our PLG funnel to drive B2B sales How we expand accounts from 3- to 6-figures
The closing slide thanks the audience and identifies Drew D’Agostino as Crystal’s CEO.
September 2022 Drew D’Agostino CEO of Crystal
