Editor’s Note: In this conversation between Alex (Corrily) and Jurn (Veed), they discuss the challenges and considerations involved when pricing AI features. Jurn shares his insights on how Veed approaches pricing for different AI functionalities and balances cost with value and competitive factors. They also touch on how to optimize monetization strategies for AI and drive activation and retention.
Jurn highlights the importance of considering the use case and value of each AI feature when deciding on pricing. While cost is a factor, it is not the sole determinant of pricing. Jurn also emphasizes the need to balance a product-led growth strategy with the costs of AI features, and how they are implementing a credit system to optimize monetization. Additionally, Jurn shares insights on optimizing activation and retention through targeted plans and free sampling motions. Overall, the interview provides valuable insights into the challenges and strategies involved in pricing AI features and monetizing AI products.
This is a follow-up to Pricing for AI Solutions and is part of a larger series of conversations happening at Corrily around monetizing AI. Want to join the conversation? Drop us a note!
Alex: I wanted to discuss the topic of integrating AI into products, specifically drawing from your recent experience with it.
Jurn: We've been integrating AI into our product and thinking about how to price it. I can provide examples from our experience.
Alex: Great, I'd love to hear about your process and learnings.
Jurn: We're still very early on in our AI journey, but we've been adding AI features like text-to-speech, avatars, and background removal. So far, each feature is priced differently based on a variety of factors like use case, cost to service, and maturity of the feature. For example, subtitles and translations look the same on the video but are used differently, so we position them differently.
Alex: That makes sense. I'm curious about how you're handling the costs of these new AI functionalities.
Jurn: We do consider the costs when pricing these features, but the cost is not leading to how we price each one. AI is also becoming a necessity in every product and we need to factor that into our monetization strategy. We set limits to ensure we are profitable, especially in our free tier, but we don't base our pricing solely on costs.
Alex: So you're considering costs when deciding where functionality will sit or how much of it you give out?
Jurn: Yes, but mostly we also consider the use case. For example, subtitles and translations have similar costs and look similar on video, but one is more useful for bigger companies as they might want to push out internal training videos or marketing campaigns in a dozen countries. That’s why we position it differently in our packaging strategy.
Alex: That's a good example of balancing cost, value, and competitive factors in pricing. AI has a real cost associated with it, unlike other features that cost only the development cost. It should be part of the pricing equation.
Jurn: Another challenge we face is balancing our PLG motion with the costs of AI features. Generally, PLG companies are generous with their free tier, to increase acquisition, activation, and engagement which will later result in more revenue overall than a more gated approach. Since AI features are quite costly, we need to make sure our PLG approach is still profitable. Especially as we have a large volume of organic SEO traffic, not scrutinizing this could cost us a lot of money.
Alex: Picking up on the PLG side of things, have you considered offering a reverse trial or credits?
Jurn: A reverse trial is difficult because many users do not have a recurring need to create videos. If we were to give these customers time-limited trials, many of them would have already gotten everything they need before we start charging them. We are implementing a credit system to set more granular limits across our AI features so we can optimize our monetization strategy for AI.
Alex: That makes sense. How are you determining feature limitations and scaling across different user bases and packages?
Jurn: We are strongly biased towards action, we ship quickly, iterate based on customer feedback and usage data, and run new experiments weekly.
Alex: Is that due to your company culture or the current fast-paced environment?
Jurn: A bit of both. We've always made decisions quickly, but the current competitive environment and investor expectations also necessitate working fast.
Alex: How are you driving larger activation and retention?
Jurn: We're targeting certain use cases with each plan and guiding users in the product based on their needs. We also had a "try before you buy" experiment that worked well for us, which allowed the use of premium features and required payment after finishing your video. We’re continuing to experiment with free sampling motions to best serve our customers.
Alex: It seems like your real monetization potential lies with the pro users. How has that impacted your packaging and pricing?
Jurn: Correct, we pay extra attention to these users when making pricing and packaging decisions. We talk to this segment often, pay attention to their usage metrics, and ask them lots of questions about the problems they run into. Solving these problems for them is the foundation that makes Veed financially sustainable.
Alex: So your packaging is more use-case oriented than a traditional good-better-best structure?
Jurn: Yes, it's a mix of both. Our good plan focuses on table stakes features, better has power features, and best is for people who collaborate with other users or have very intense usage patterns.
Alex: That's interesting. It seems like the value is in the content created, so perhaps pricing could be based on the amount of content produced.
Jurn: We're exploring options, but how many videos someone creates doesn't necessarily correlate with how much they're willing to pay.
Alex: That's a tricky problem. Any other challenges with AI recently?
Jurn: We just launched AI avatars, which could be difficult because companies might want to bring in their own data, which is sensitive. We might need to support on-premise or have different enterprise conversations, but those aren't all related to pricing.
Alex: That's a good point. It's important to consider data sensitivity when integrating AI. Well, it was great discussing this with you, Jurn. Let's keep in touch as we continue to explore these issues.
