ChatGPT
Claude
Google AI
Copilot
A programme lead finds a 2019 acquisition and asks where the product went, because the migration template needs a destination. The template has a source column and a target column. Both columns are blank, and the lead is about to type a Salesforce product name into the target so the row looks finished. Do not migrate onto a product this page cannot name. No current Bonobo screen is established. Filling the blank is how a rumour becomes a project.
This page will not repeat a reported dollar figure. A number in a headline is not a licence, and it is not a budget for your move. The figure stays out of the briefing, the footnote, and the FAQ.
Should you migrate
No, not onto a named Salesforce product, because this page cannot name one honestly. A migration needs a system you are leaving and a system you are joining. The report this page cites tells you what Bonobo developed. It does not tell you which screen a seller opens in 2026 under that name. Without that screen, a cutover is a story.
If your company never ran Bonobo, there is nothing of yours to move. The acquisition is a fact about Salesforce and a company it bought. It is not a tenant in your estate. Do not invent a historical extract so the project has files. Absence is a finding. Write "we did not run this" and close the line.
If someone on the team remembers a pilot, treat the memory as a prompt to look for a contract, a login, or an export, not as a map. A remembered demo is not a system of record. Until the export is in front of you, you do not have data to migrate. You have a recollection. Recollections do not get field mappings.
The walkthrough is therefore a search for evidence, and a stop when the evidence is missing. Look for the contract. Look for the admin. Look for an export format you can open. If all three are missing, stop. Do not continue into a target system you picked because the template required one. The stop is the responsible outcome, not a failed project.
What Bonobo is now
The sentence this page quotes from the report is: "Bonobo develops conversational AI technology designed to extract insights from online customer interactions." Develops, in that sentence, is what the company did. Conversational AI, in that sentence, means technology aimed at insights from interactions customers had online. That is the description this source supports.
Salesforce bought Bonobo in 2019. The report this page cites is dated 6 May 2019. That date is the date of the report. It is not a go-live in your org, and it is not a sunset. This page does not add a second date for a rename or a shutdown. Those dates are not in the summary this guide is willing to stand on.
No current Bonobo screen is established. Read that as a limit on this page, not as a hint that you should go looking for a secret menu. If a current product exists under another name, this source does not identify it. Naming one from memory would be a guess. The guess would then appear in a design document, and people would migrate towards it. That is the failure mode. Stop at the description in the quotation.
What the software is now, for a team deciding a move, is an acquired technology from 2019 with no screen this page can point to. Treat it as history. Do not treat it as a SKU on a renewal. Do not treat it as a connector you can install. Installation implies a package. This page does not have a package.
The reported price of the deal is a different hunger. This page does not repeat it. It does not round it, range it, or say "about" it. A licence fee is not established here either. Your contract is the only commercial paper that would matter, and this guide is not your contract. If finance asks what Salesforce paid, the answer from this page is that the figure is not being used. Send them to a source you are willing to cite in full, or leave the cell blank.
Do not name a landing zone
The tempting names are the products you already own. Service Cloud, because the words were customer interactions. A chatbot you already configured, because the words were conversational. An analytics object, because the words were insights. Each of those may be a real system in your org. None of them is identified by this source as the place Bonobo went. Using them as the destination borrows your org to complete a sentence the report left unfinished.
Write the refusal on the diagram. "Destination: not named. This page will not name one." If a reviewer insists on a box, the box says unknown, and the arrow stops. An arrow that continues into a product you like is a design decision you have not earned. Make that decision under the product's own name, with its own requirements, without the 2019 brand as a mascot.
Insights from customer interactions, as a job, can still be a job you want. Scope it as the job. Which interactions. Which customers. Who may read the text. What is stored. How long it is kept. Those questions belong to whatever tool you already use to hold conversations. They do not belong to a migration from a product you cannot open.
Do not let a partner's proposal fill the destination in a footnote. Footnotes are where guessed product names go to look official. If the footnote names a product, ask which sentence on the cited report identifies it. If they cannot answer, delete the footnote. The proposal can still sell you help. It cannot sell you a lineage this page does not support.
Versus native features
Salesforce already stores customer interactions when you use the service tools in your contract. Cases, chat transcripts, emails on a record, and the reports you built on them are native to the org that holds them. They exist whether or not a company called Bonobo was bought in 2019. They do not become a migration target by being nearby in vocabulary.
The report describes technology designed to extract insights from those kinds of interactions. Extracting an insight is a different job from storing the interaction. Native records keep what was said, under the permissions on the record. An insight layer, if you build one, needs its own rules about what it may read and what it may write back. This page will not claim that Bonobo is that layer in your org today.
This page does not rank a hand-built report against an acquired technology it cannot show you. There is nothing to rank. The native record is something you can open. The Bonobo screen is not established. Comparison, in that situation, is a check that you are not about to replace a system you can see with a system you cannot. Keep the system you can see until a current product is named by a source you trust, or by your own contract.
| Job | What this record says about Bonobo | What native features already do | What you still write |
|---|---|---|---|
| Store a customer conversation | Not described as your system of record | Cases and related messages, if you use them | The channel you actually run |
| Extract an insight from that conversation | The 2019 description of the technology | A report on records you already have | The question, the fields, and who may see the result |
| Move off a Bonobo tenant | No current screen is established here | Nothing migrates itself | Evidence of a tenant, or a decision to stop |
| Land on a named Salesforce product | This page cannot name one | Your contract lists the products you have | Do not type a guess into the target column |
The last row is the whole decision. Leave the target blank. A blank target is allowed to end the project.
Alternatives, with no order
An AppExchange search wants a listing that replaces Bonobo. This page does not name a listing, a publisher, or a rank. Naming one would invent a directory entry and, worse, would name the landing zone the previous section forbids. The options are ways to handle the job. They are not ordered.
Keep the conversation records in the system that already holds them, and report on those records with tools you already licence. That option fits when the need is "understand what customers said" and the text is already in the org. You do not need a 2019 brand to open a report.
Scope a new insight question against a product your contract names, with fields and permissions written down. That option fits when a report is not enough and you are willing to design the job properly. The design uses the product's name. It does not use Bonobo as a synonym.
Look for evidence you actually ran Bonobo, and stop if you find none. That option fits almost every team that arrived here from a search. The search is not a tenancy.
Do nothing. There is no renewal rushing you, because no licence is established on this page, and no shutdown date is established either. Inventing an end date so the migration feels urgent would be another fiction. This page will not supply one.
A partner who ranks alternatives, or who puts a price from 2019 into the commercial section, is mixing a headline with your project. Ask them to remove both the rank and the figure. If the proposal still makes sense as a scoped piece of work on a product you can name, you can read it on those terms. If it only makes sense as a Bonobo migration, it does not make sense.
What the walkthrough contains
The walkthrough has gates, and most teams stop at gate two.
Gate one. Write "Salesforce bought Bonobo in 2019" and the description from the report. Do not add a dollar figure beside it.
Gate two. Search your contracts, identity provider, and inboxes for a tenant. If you find nothing, stop. Record the stop. Do not open a sandbox "to receive the migration".
Gate three, only if gate two found evidence. List what the export contains. Interactions, insights, users, or something else. Each type needs an owner. If the export is empty, stop. An empty export is not a prompt to synthesise rows.
Gate four. Do not name a target product on this page's authority. If you choose a target, the choice cites your contract or a current product page you have actually read, and the project is renamed to that product. The Bonobo name comes off the title.
Gate five. Permissions and retention for any text you did find. Who may read a customer interaction. How long you keep it. A migration that copies text into a wider audience is not a completion. It is a disclosure.
There is no implementation cost on this page. There is no ranking of agencies. A quote that arrives with both has answered a search, not your estate.
Where the other pages fit
The implementation page is the place for what 2019 does and does not put on an order form: Bonobo, conversation insight, and what 2019 does not sell you. Go there when someone is still trying to buy it. Do not go there for a screen this source does not establish.
What happened, the year, and the refusal to repeat a reported figure are on what happened to Bonobo AI. Stay there when the question is the acquisition as history. Leave it when the question is whether you have data to move.
The integration page is the diagram with the empty arrow: integrating Bonobo AI with Salesforce. Read it before anyone draws a connector. A connector needs two ends. This page can name the 2019 description. It cannot name the current end.
| Question | Where it belongs |
|---|---|
| Should we migrate? | This page. Not onto a product this page cannot name. |
| What is the software now? | A 2019 description. No current screen is established. |
| Versus native? | Conversation records you already store are a different job. No winner. |
| Alternatives? | Report on what you have, scope a named product, look for evidence, or stop. |
The row in the tracker can end this week. Source unknown or not in our estate. Target not named. Decision: do not migrate. That is a complete status. It does not need a phase two. If a later contract names a current product and a real export, that is a new project with a new title. It is not a revival of this one.
