B2B AI Chatbot SaaS
Thinkstack had a truth simple enough to be missed.
Thinkstack had a truth simple enough to be missed.
Founded in 2023, Thinkstack entered the market at the moment automated reply bots were giving way to AI-native products built on GPT models.
The product was early-stage with no sharply defined ICP. A handful of blogs existed, all broad category-level education. Traffic sat in the low hundreds, and the searches were pulling in readers rather than buyers. For a product this easy to start with, uptake was low.
The request was to position Thinkstack to stand up against its competitors. The read pointed somewhere else entirely.
The product was early-stage with no sharply defined ICP. A handful of blogs existed, all broad category-level education. Traffic sat in the low hundreds, and the searches were pulling in readers rather than buyers. For a product this easy to start with, uptake was low.
The request was to position Thinkstack to stand up against its competitors. The read pointed somewhere else entirely.
Thinkstack had a truth simple enough to be missed.
B2B AI Chatbot SaaS
Founded in 2023, Thinkstack entered the market at the moment automated reply bots were giving way to AI-native products built on GPT models.
Thinkstack had a truth simple enough to be missed.
Founded in 2023, Thinkstack entered the market at the moment automated reply bots were giving way to AI-native products built on GPT models.
B2B SaaS · AI Chatbot
What the evidence contradicted
What the evidence contradicted
What the evidence contradicted
What the evidence contradicted
01
The product had exactly one real characteristic. It was a blank slate that became whatever you trained it on. That is a plain sentence. It was also the only claim that was completely true, and the only one the product could carry without qualification.
02
The free tier was materially better than the category's and nobody knew. Training credits on the free tier were set higher than what competitors were offering.
Read more
03
Category competitors owned complex integration use cases, enterprise claims, and vertical-specific multichannel deployments. All of them were racing to claim the most advanced position before it solidified.
Read more
04
The free AI chatbot space was competitive, with Zapier and HubSpot leading. But buyers were searching generator and creator variations that no one had taken.
04
The free AI chatbot space was competitive, with Zapier and HubSpot leading. But buyers were searching generator and creator variations that no one had taken.
The judgment this required
The judgment this required
The judgment this required
The judgment this required
The decision to position downward while the entire category raced upward is not in that evidence. The internal pushback was immediate and reasonable. The team agreed the existing homepage was already good, and that plainness would make Thinkstack less competitive in a market full of players claiming more.
The decision to position downward while the entire category raced upward is not in that evidence. The internal pushback was immediate and reasonable. The team agreed the existing homepage was already good, and that plainness would make Thinkstack less competitive in a market full of players claiming more.
The decision to position downward while the entire category raced upward is not in that evidence.
The internal pushback was immediate and reasonable. The team agreed the existing homepage was already good, and that plainness would make Thinkstack less competitive in a market full of players claiming more.
The claim that held
The claim that held
The claim that held
The claim that held
The smallest SMB buyer — no IT team, no budget for custom development — needs a custom AI chatbot trained on their own data, and it being free to start is the whole reason they move.
Two decisions followed:
The smallest SMB buyer — no IT team, no budget for custom development — needs a custom AI chatbot trained on their own data, and it being free to start is the whole reason they move.
Two decisions followed:
Position for the most honest use case and the buyer it actually fit, rather than the buyer the category was competing over.
Position for the most honest use case and the buyer it actually fit, rather than the buyer the category was competing over.
Build the organic engine around an unclaimed keyword that matched the freemium model exactly.
Build the organic engine around an unclaimed keyword that matched the freemium model exactly.
What the claim then organised
What the claim then organised
What the claim then organised
What the claim then organised
01
The keyword stopped being the product's own name, which nobody searched for, and became "free AI chatbot generator."
01
The keyword stopped being the product's own name, which nobody searched for, and became "free AI chatbot generator."
02
The homepage dropped the efficiency-and-toolkit vocabulary every competitor already owned, and put the free tier above the fold.
02
The homepage dropped the efficiency-and-toolkit vocabulary every competitor already owned, and put the free tier above the fold.
03
Alternatives pages went up for the tools this buyer was already paying for and wanted to leave.
03
Alternatives pages went up for the tools this buyer was already paying for and wanted to leave.
04
A three-layer funnel went in behind all of it, so arrivals from any entry point landed on the same conversion path.
04
A three-layer funnel went in behind all of it, so arrivals from any entry point landed on the same conversion path.
THEN THE PRODUCT OUTGREW WHAT THE CLAIM DESCRIBED
THEN THE PRODUCT OUTGREW WHAT THE CLAIM DESCRIBED
Five months in, new integrations were live and the product sat between a standard chatbot and a full autonomous agent — two categories, with no page speaking either buyer's language. The blank-slate claim still held but what it described had grown.
Five months in, new integrations were live and the product sat between a standard chatbot and a full autonomous agent — two categories, with no page speaking either buyer's language. The blank-slate claim still held but what it described had grown.
Homepage sections surfaced agentic language only where the product supported it.
Homepage sections surfaced agentic language only where the product supported it.
Use case pages split into two sets, chatbot and AI agent, each with its own buyer and conversion path, rather than forcing the product into one category and making it smaller than it was.
Use case pages split into two sets, chatbot and AI agent, each with its own buyer and conversion path, rather than forcing the product into one category and making it smaller than it was.
A non-indexed sales page became the enterprise leave-behind, built around what the buyer was weighing internally rather than a deck repeating the demo.
A non-indexed sales page became the enterprise leave-behind, built around what the buyer was weighing internally rather than a deck repeating the demo.
The founders' origin became a page in the demo-confirmation sequence, addressing what prospects kept asking about directly.
The founders' origin became a page in the demo-confirmation sequence, addressing what prospects kept asking about directly.
THEN THE PRODUCT OUTGREW WHAT THE CLAIM DESCRIBED
Five months in, new integrations were live and the product sat between a standard chatbot and a full autonomous agent — two categories, with no page speaking either buyer's language. The blank-slate claim still held but what it described had grown.
Homepage sections surfaced agentic language only where the product supported it.
Use case pages split into two sets, chatbot and AI agent, each with its own buyer and conversion path, rather than forcing the product into one category and making it smaller than it was.
A non-indexed sales page became the enterprise leave-behind, built around what the buyer was weighing internally rather than a deck repeating the demo.
The founders' origin became a page in the demo-confirmation sequence, addressing what prospects kept asking about directly.
What changed
What changed
What changed
What changed
Ten months of building followed, through a product that changed underneath it. The claim wasn't revised, and no surface built on it had to be walked back — the architecture extended around the growth rather than replacing it. It kept converting the same way, deal after deal, without anyone returning to argue for it.
Ten months of building followed, through a product that changed underneath it. The claim wasn't revised, and no surface built on it had to be walked back — the architecture extended around the growth rather than replacing it. It kept converting the same way, deal after deal, without anyone returning to argue for it.
Ten months of building followed, through a product that changed underneath it. The claim wasn't revised, and no surface built on it had to be walked back — the architecture extended around the growth rather than replacing it.
It kept converting the same way, deal after deal, without anyone returning to argue for it.
~12x traffic growth. ~90x lead growth. Ten months (Aug 2024 - May 2025).
Read more
Every decision was grounded in research and data, never assumptions...started with understanding the product before changing the narrative. The result was clear, market-facing communication that directly supported the sales motion.
Rajesh Bhattacharjee
CEO & Co-Founder @Pickcel, Thinkstack
The content and messaging brought clarity to our prospect conversations, making it easier to communicate the value clearly.
Akash Ranjan
AVP Business Development @Pickcel
There was a clear positioning focus in how both brands were communicated. The messaging stayed consistent while being adapted to each product and audience. It gave the team clearer direction and sharpened the quality of what was delivered.
Deblina Chatterjee
Senior HR Manager @Pickcel, Thinkstack
High ownership of the product narrative and the ability to turn complex ideas into clear story…..Strong research grounded in the customer’s perspective made the technical products easier to communicate and gave teams clearer direction across functions.
Nilotpal Boruah
Co-founder & CEO @TechVariable
Every decision was grounded in research and data, never assumptions...started with understanding the product before changing the narrative. The result was clear, market-facing communication that directly supported the sales motion.
Rajesh Bhattacharjee
CEO & Co-Founder @Pickcel, Thinkstack
The content and messaging brought clarity to our prospect conversations, making it easier to communicate the value clearly.
Akash Ranjan
AVP Business Development @Pickcel
There was a clear positioning focus in how both brands were communicated. The messaging stayed consistent while being adapted to each product and audience. It gave the team clearer direction and sharpened the quality of what was delivered.
Deblina Chatterjee
Senior HR Manager @Pickcel, Thinkstack
High ownership of the product narrative and the ability to turn complex ideas into clear story…..Strong research grounded in the customer’s perspective made the technical products easier to communicate and gave teams clearer direction across functions.
Nilotpal Boruah
Co-founder & CEO @TechVariable
Every decision was grounded in research and data, never assumptions...started with understanding the product before changing the narrative. The result was clear, market-facing communication that directly supported the sales motion.
Rajesh Bhattacharjee
CEO & Co-Founder @Pickcel, Thinkstack
The content and messaging brought clarity to our prospect conversations, making it easier to communicate the value clearly.
Akash Ranjan
AVP Business Development @Pickcel
There was a clear positioning focus in how both brands were communicated. The messaging stayed consistent while being adapted to each product and audience. It gave the team clearer direction and sharpened the quality of what was delivered.
Deblina Chatterjee
Senior HR Manager @Pickcel, Thinkstack
High ownership of the product narrative and the ability to turn complex ideas into clear story…..Strong research grounded in the customer’s perspective made the technical products easier to communicate and gave teams clearer direction across functions.
Nilotpal Boruah
Co-founder & CEO @TechVariable
About this engagement
A note on results
Recent Works
Recent Works
Recent Works
The diagnostic archive
The diagnostic archive
The diagnostic archive
The diagnostic archive
Selected work led by us before Legibia.
Selected work led by us before Legibia.
Selected work led by us before Legibia.
Selected work led by us before Legibia.
THEN THE PRODUCT OUTGREW WHAT THE CLAIM DESCRIBED
THEN THE PRODUCT OUTGREW WHAT THE CLAIM DESCRIBED
Five months in, new integrations were live and the product sat between a standard chatbot and a full autonomous agent — two categories, with no page speaking either buyer's language. The blank-slate claim still held but what it described had grown.
Homepage sections surfaced agentic language only where the product supported it.
Use case pages split into two sets, chatbot and AI agent, each with its own buyer and conversion path, rather than forcing the product into one category and making it smaller than it was.
A non-indexed sales page became the enterprise leave-behind, built around what the buyer was weighing internally rather than a deck repeating the demo.
The founders' origin became a page in the demo-confirmation sequence, addressing what prospects kept asking about directly.