Most negotiations on WhatsApp do not fail because the deal was bad. They fail because nobody tracked the state of the deal.
A WhatsApp negotiation looks like this: an offer is made on Monday. A counteroffer comes back on Wednesday, but in a voice message while the other person is driving. On Friday, someone says "ok" to something, but it is unclear which version they are agreeing to. By the following week, each side remembers a different deal.
The gap between "what was discussed" and "what was agreed" is where deals fall apart and disputes begin.
Turn this thread into meeting notes and action items.
Analyze your chatA negotiation recap closes that gap. It extracts the current state of a negotiation from a conversation that was never designed to be structured.
A recap is not a transcript. It is a structured summary that answers:
WhatsApp threads are non-linear in practice. A conversation that started as a simple price discussion can drift into delivery dates, then back to price, then into payment terms, all within the same thread. When you read the thread top to bottom, it can be difficult to isolate the final position on any single term because it is scattered across dozens of messages and days.
A chronological offer log solves this directly. By listing each offer with a date and the specific term it applies to, you can see the arc of the negotiation at a glance: who moved first, how much ground was given up, and at what point the conversation stalled. That sequence is also what you need if the deal later becomes a dispute. A timestamp on a WhatsApp message is metadata you can point to. "You agreed on the 14th" is a claim. "Your message at 11:42 on the 14th said X" is evidence.
Most negotiations do not resolve at the headline number. They resolve through a series of smaller concessions, each of which shifts the deal slightly. Tracking those concessions separately from the main offer log is worth the effort because it surfaces patterns. If one side has made three concessions and the other has made none, that is a leverage imbalance that should inform your next move. If a concession was made without getting anything in return, that is worth noting before you give up another one.
[One paragraph: where the negotiation stands right now. What is agreed, what is not, what is the main blocker.]
For each concession:
A ThreadRecap analysis is designed for exactly this. It extracts positions, arguments, leverage points, red lines, and suggests next steps.
Here is the full workflow:
Export the conversation and upload it to ThreadRecap. Include media if voice messages contain key terms (the real offers are often in voice messages, not text).
Run the analysis on the uploaded export. It produces:
This single analysis often gives you 80% of what you need. Cost: 1 credit per 1,000 messages, plus 1 credit per 10 minutes of audio, so a typical negotiation thread lands at a credit or two (roughly $0.30-0.50 depending on your pack).
For negotiations involving specific terms (price, payment timing, delivery, penalties), ask in the chat that opens with the recap:
Build a checklist of deal terms mentioned in this conversation: price, payment timing, delivery date, scope, warranty, penalties, cancellation, dependencies. For each term, show what was said by each side and whether it was agreed, disputed, or unclear.
This gives you a term-by-term view that complements the narrative recap.
Voice messages are a particular problem in WhatsApp negotiations. They are fast to send, so people use them for substantive statements, but they leave no searchable text record. A supplier might state a revised delivery date in a voice message, the buyer says "ok" in the next text message, and there is now a binding commitment with no written evidence of what the "ok" refers to.
ThreadRecap transcribes WhatsApp voice messages in .opus and .m4a formats using OpenAI's transcription models. On clear audio the transcription is reliable enough to pull specific terms out of voice messages and include them in the offer log. When you upload a WhatsApp export with the media folder included, those voice message files are processed alongside the text, and the transcript is used as part of the analysis. If a price was quoted verbally on a Tuesday morning voice message, it will appear in the output alongside the text messages that followed it.
The most valuable output of a negotiation recap is not the document itself. It is the confirmation message you send afterward. A message that says "here is what I understand we agreed" creates a written record and forces the other side to either confirm or correct.
Template:
Hi [Name],
To make sure we are aligned, here is where I think we stand:
Agreed:
Still open:
Next steps:
If anything above is not accurate, let me know and we will adjust. Otherwise I will move forward on this basis.
Thanks,
[Your name]
Do not wait until the negotiation is over. Build recaps at these moments:
If a negotiation has been running for more than a week, the volume of messages alone becomes a problem. ThreadRecap does not set a fixed message count and accepts ZIP files up to 2 GB, so extended commercial negotiations with dense back-and-forth can be uploaded for review. But size is not the only challenge: in a long thread, early positions get lost, people forget what they already conceded, and conversations that were exploratory at the start get treated as binding later.
Running a Negotiation Analysis mid-process creates a checkpoint. You get a clear statement of where things stand at that moment, which you can use to anchor the next phase of the discussion. If the other side tries to walk back a concession they made three weeks ago, you have a dated record from the ThreadRecap analysis that captured it at the time.
Negotiations are sensitive. You probably do not want your deal terms stored on a random server. ThreadRecap processes exports client-side first: the file is unzipped and previewed in your browser before any data is sent. Photos never leave your device, and video files are never uploaded; if you include videos, your browser extracts their audio track and sends only that. Only the chat text (and voice message audio, if you opt in) are sent for analysis. Read the full privacy details.
When you drag a WhatsApp export ZIP into ThreadRecap, the file is decompressed and read entirely within your browser tab before any network request is made. You can see the message count, date range, and participant list before you commit to uploading. If the export contains photos or videos, those files are identified and kept out of the upload automatically. They never leave the device; the copy your browser holds so an interrupted upload can resume is gone once you clear the browser's data. Including videos in the analysis sends the audio track your browser extracts from them, never the file itself. This matters for commercial negotiations where the thread might include contract drafts, product images, or other sensitive attachments that are not relevant to the analysis anyway.
The text content and any voice message audio you opt into are sent over an encrypted connection for analysis. ThreadRecap uses OpenAI and Claude models depending on the analysis type. The data is processed for the purpose of generating your analysis and is not used for model training. For most business users, this is a material difference from pasting a conversation into a general-purpose AI chat interface where retention and use policies are less clear.
Upload your WhatsApp export to ThreadRecap and run the analysis to see exactly where things stand. 5 free credits on signup. Pay-as-you-go credits start at $2.20, purchased credits never expire, no subscription.
Upload your export and get decisions, action items, and a clean summary you can forward in minutes.
Turn WhatsApp client chats into a running change log that tracks scope, approvals, and costs so you can stop delivering free work and invoice with confidence.
Feb 10, 20268 min read
Create a WhatsApp negotiation recap that tracks offers, counteroffers, and agreements clearly, preventing miscommunication and deal disputes from the start.