I’ve been looking into $TEXT.WA, and despite my tech background, I couldn’t find any MOAT for this business. What do you think is their competitive advantage? Their app ecosystem seems pretty good, but the switching costs aren’t that significant.
I'd say that TEXT's competitive advantage lies in its integration and ecosystem stickiness. The switching costs may not be immediately obvious.
Integrated Suite, marketplace and API monetization will increase switching costs especially for enterprise customers. All this create custom solutions tailored to clients's needs, making it harder for them to switch providers.
Also, keep in my mind TEXT's own language model. The company has been in the business for several year now, serving thousends of customers. Think how much data they can process now and learn from. Data is the new oil.
If you can't see competitive adventages, what other players are better than Text and could serve big customers?
- Under the assumption of losing 5-8% of customers quarterly, how do you reach an annual reduction of 15.5%? The range would be 18.5-28.5% per year.
- How does 3 years of losing 15.5% of customers per year with 5% ARPU growth get to third-year revenue decline of only 14%? I'm getting 25%. (If you used 18.5-28.5%, the decline is instead 30-45% by my calculation)
- Why would a company losing revenue this quickly as well as margins dropping trade at 10x EBIT? Being a software company doesn't do you any favours when your financials look like that - just look at Chegg
- If AI can do what they can do better, why will this company exist at all in 10 years' time?
Customer losses has its reason. Quarterly churn is partially offset by new customer acquisition. The company reguraly gets more clients from different industries. They are doing strategic customer shift to higher-value clients.
The revenue decline assumes partial replacement through new enterprise clients and higher monetization of API/HelpDesk. I can't ignore the latest reports. Even using more conservative approach, current valuation remains compellig.
Yeah, I looked at Chegg. The company had 10% of EBIT margins at the highest during the C19 bubble and many years with negative margins. The company was loaded with debt and was issuing shares. TEXT's fundamentals remain exceptional compared to Chegg.
And finally, Text is not competing with AI. It is leveraging it as a strategic tool. Text has been using AI technology for years.
Aside from the last point, I was only basing my comments on things that you said yourself - my intention was to show they weren't internally consistent, rather than that this stock is bad.
Imo the fact that they are also getting new clients is not relevant, since we're already working with net loss of customer numbers, not gross.
The strategic shift to higher-value clients is relevant though. I would say just make sure that's genuinely what's going on, rather than an excuse.
On the AI point - ChatGPT flipped AI on its head in 2023. Any model they had prior to that became worth very little at that point. If they're genuinely managing to stay right at the forefront in implementing these models into their live chat, they may do very well. But I'd be cautious that the chatbot space has become incredibly democratised in the last couple of years, and they may be facing much stiffer competition now.
But I don't know this company so take with a huge pinch of salt
Agree with you. We cannot forget that founders (main shareholders) are also computer engineers and well aware of the risks and opportunities of AI. I think that at these prices the annual return will be attractive for shareholders with a long term time horizon.
Why doesn't AI just straight up kill this business?
Core Livechat seems like it will be replaced by AI (see Klarna sacking lots of their workforce and replacing with AI chatbots). Text's AI offerings are likely not competitive with mega cap LLMs (or even if they are competitive, they now have far more competitors to compete against). Numerous companies have come out with their own chatbots.
I think your nightmare scenario isn't aggressive enough. Why does ARPU keep growing? Why doesn't it shrink with price competition? Churn is already ~4% a month and competition is only increasing. Why is customer count not falling more than 15%? 15% implies a lot of gross additions/year to stay above the churn rate.
35% EBIT margins in your nightmare case is higher than a great deal of software businesses in less competitive markets.
I believe Mr. Market has already priced extreme pessimism.
Text isn't fighting AI but integrating it. E.g. ChatBot adapt solutions and businesses will rather adopt hybrid human-AI solutions, and not eliminate humans entirely. The Klarna example isn't directly comparable to Text. Text is a platform provider, not just a service provider. And they are also diversified among different industries.
ARPU has increased despite competitors. It looks like their strategy works. The churn primarily affects smaller customers. The company is moving toward clients with lower churn rates and higher lifetime value.
Even if you assume: 30% customer decline, margins compressed to 25%, zero ARPU growth, you may land on 12-13x EBIT.
If the stock market prices in the future 9-12 months ahead, how bad would the scenario have to be to justify the current price?
LLM will not replace cases where human action is required. Unless we have robots. Other cases are where human is required for security reasons. Text will rather benefit from LLM inside it.
I'm sure we will still need human agents in the future. But there will be fewer of them (decreasing clients) and pricing will be lower (as the same number of companies compete for fewer clients).
And then how does Text benefit from using LLMs when they compete against the owners of said LLMs?
Does it mean less agent seats ? Hard to tell. Pricing, hard to tell :-) Same number of human agents will have more work done. Amount of work is likely increase as digitalization taking over the world. So many countries will join the process.
Text have everything to benefit from LLMs. You want an AI agent to have at least some idea of the subject, so you need to train it on the customer data. Some data might be sensitive - money, personal data, company secrects so companies will want to choose trustworthy partner to do so. Some would prefer this training data stored within iown servers. So it is more to it then just simple chatbot replacing human. LLM owners not going to build this, they have their own moonshots and most likely live chat is not one of them.
I’ve been looking into $TEXT.WA, and despite my tech background, I couldn’t find any MOAT for this business. What do you think is their competitive advantage? Their app ecosystem seems pretty good, but the switching costs aren’t that significant.
Switching costs for these kind of business are huge. Coming from an ops background. We are stuck due to integrations
Hi Alexandru.
I'd say that TEXT's competitive advantage lies in its integration and ecosystem stickiness. The switching costs may not be immediately obvious.
Integrated Suite, marketplace and API monetization will increase switching costs especially for enterprise customers. All this create custom solutions tailored to clients's needs, making it harder for them to switch providers.
Also, keep in my mind TEXT's own language model. The company has been in the business for several year now, serving thousends of customers. Think how much data they can process now and learn from. Data is the new oil.
If you can't see competitive adventages, what other players are better than Text and could serve big customers?
A few things.
- Under the assumption of losing 5-8% of customers quarterly, how do you reach an annual reduction of 15.5%? The range would be 18.5-28.5% per year.
- How does 3 years of losing 15.5% of customers per year with 5% ARPU growth get to third-year revenue decline of only 14%? I'm getting 25%. (If you used 18.5-28.5%, the decline is instead 30-45% by my calculation)
- Why would a company losing revenue this quickly as well as margins dropping trade at 10x EBIT? Being a software company doesn't do you any favours when your financials look like that - just look at Chegg
- If AI can do what they can do better, why will this company exist at all in 10 years' time?
Ai doesnt do what they do. They do the website integrations. You could use text.wa with 50 chatbots or 50 agents, no different for text.
Customer losses has its reason. Quarterly churn is partially offset by new customer acquisition. The company reguraly gets more clients from different industries. They are doing strategic customer shift to higher-value clients.
The revenue decline assumes partial replacement through new enterprise clients and higher monetization of API/HelpDesk. I can't ignore the latest reports. Even using more conservative approach, current valuation remains compellig.
Yeah, I looked at Chegg. The company had 10% of EBIT margins at the highest during the C19 bubble and many years with negative margins. The company was loaded with debt and was issuing shares. TEXT's fundamentals remain exceptional compared to Chegg.
And finally, Text is not competing with AI. It is leveraging it as a strategic tool. Text has been using AI technology for years.
Aside from the last point, I was only basing my comments on things that you said yourself - my intention was to show they weren't internally consistent, rather than that this stock is bad.
Imo the fact that they are also getting new clients is not relevant, since we're already working with net loss of customer numbers, not gross.
The strategic shift to higher-value clients is relevant though. I would say just make sure that's genuinely what's going on, rather than an excuse.
On the AI point - ChatGPT flipped AI on its head in 2023. Any model they had prior to that became worth very little at that point. If they're genuinely managing to stay right at the forefront in implementing these models into their live chat, they may do very well. But I'd be cautious that the chatbot space has become incredibly democratised in the last couple of years, and they may be facing much stiffer competition now.
But I don't know this company so take with a huge pinch of salt
Agree with you. We cannot forget that founders (main shareholders) are also computer engineers and well aware of the risks and opportunities of AI. I think that at these prices the annual return will be attractive for shareholders with a long term time horizon.
Thanks for the writeup.
Why doesn't AI just straight up kill this business?
Core Livechat seems like it will be replaced by AI (see Klarna sacking lots of their workforce and replacing with AI chatbots). Text's AI offerings are likely not competitive with mega cap LLMs (or even if they are competitive, they now have far more competitors to compete against). Numerous companies have come out with their own chatbots.
I think your nightmare scenario isn't aggressive enough. Why does ARPU keep growing? Why doesn't it shrink with price competition? Churn is already ~4% a month and competition is only increasing. Why is customer count not falling more than 15%? 15% implies a lot of gross additions/year to stay above the churn rate.
35% EBIT margins in your nightmare case is higher than a great deal of software businesses in less competitive markets.
I believe Mr. Market has already priced extreme pessimism.
Text isn't fighting AI but integrating it. E.g. ChatBot adapt solutions and businesses will rather adopt hybrid human-AI solutions, and not eliminate humans entirely. The Klarna example isn't directly comparable to Text. Text is a platform provider, not just a service provider. And they are also diversified among different industries.
ARPU has increased despite competitors. It looks like their strategy works. The churn primarily affects smaller customers. The company is moving toward clients with lower churn rates and higher lifetime value.
Even if you assume: 30% customer decline, margins compressed to 25%, zero ARPU growth, you may land on 12-13x EBIT.
If the stock market prices in the future 9-12 months ahead, how bad would the scenario have to be to justify the current price?
It doesn’t make sense.
LLM will not replace cases where human action is required. Unless we have robots. Other cases are where human is required for security reasons. Text will rather benefit from LLM inside it.
Agentic AI is already replacing needs for humans:
https://www.bbc.co.uk/news/articles/c80e1gp9m9zo
I'm sure we will still need human agents in the future. But there will be fewer of them (decreasing clients) and pricing will be lower (as the same number of companies compete for fewer clients).
And then how does Text benefit from using LLMs when they compete against the owners of said LLMs?
E.g. https://www.geo.tv/latest/593839-whatsapps-new-meta-ai-upgrade-will-improve-user-chatbot-interactions
Text still gets paid if you use an AI agent
Does it mean less agent seats ? Hard to tell. Pricing, hard to tell :-) Same number of human agents will have more work done. Amount of work is likely increase as digitalization taking over the world. So many countries will join the process.
Text have everything to benefit from LLMs. You want an AI agent to have at least some idea of the subject, so you need to train it on the customer data. Some data might be sensitive - money, personal data, company secrects so companies will want to choose trustworthy partner to do so. Some would prefer this training data stored within iown servers. So it is more to it then just simple chatbot replacing human. LLM owners not going to build this, they have their own moonshots and most likely live chat is not one of them.