Customer success managers prep QBRs faster when AI does the gathering, not the thinking. It pulls usage, tickets, call notes and open actions from the CRM into one draft, with a source next to every number. The CSM checks the figures, decides what the account needs to hear, and writes the recommendations themselves.
#Key takeaways
- Most QBR prep time goes on collecting data from five or six tools, which is the part AI handles well.
- The story, the risks and the asks for the customer should stay with the CSM, because that is the judgement the customer pays attention to.
- Every number in an ai-built draft needs a link back to the system it came from, or nobody will trust the deck.
- A shared template and a CRM export are enough for a small book of accounts; a built workflow pays off when many CSMs prep many reviews.
- Measure prep effort per review and the share of reviews that end with agreed next steps, before and after.
#Why does QBR prep eat so much of a CSM's week?
A typical CSM at a B2B services or software company runs a book of accounts across Salesforce or HubSpot, Zendesk or Freshdesk, a product usage dashboard, Gong or Teams call recordings and a folder of old decks. Before each review they copy numbers out of each tool, chase the account executive for commercial context and rebuild last time's slides.
This is not a niche complaint. Bain, december 2025 reports that customer success managers spend about two-thirds of their time on lower-value work that could be automated with AI. QBR assembly is a large share of that work, and it leaves little room to think about what the customer should do next. Teams are already reaching for help: a Gainsight customer success index release, january 2025 found that 52% of CS teams already use AI.
#What should AI do in QBR prep, and what stays with the CSM?
- AI gathers: usage trends, ticket volumes and themes, open escalations, actions promised last review, and key quotes from call notes.
- AI drafts: a first version of the account summary and the 'what happened' slides, each figure linked to its source.
- AI flags: gaps such as a missing executive sponsor, a feature bought but unused, or an action that was never closed.
- the CSM decides: the narrative, the risks to raise, the recommendations, and what to ask the customer for.
- the CSM checks: every number against the source before the deck reaches the customer or the account executive.
The split matters because a QBR is a trust exercise. A customer who spots one wrong figure starts doubting the rest. Keep AI on the reading and the first draft, and keep the person who owns the relationship on the message.
#A QBR prep playbook you can use this cycle
- Agree one QBR template across the team: goals agreed last time, results against them, support picture, adoption, risks, next steps.
- List the systems each section draws from and who owns each one.
- Export the account data and paste it, with last review's notes, into an approved AI tool such as ChatGPT Enterprise, Copilot or Claude.
- Ask for a draft per section with the source of every figure, and ask what is missing.
- Check each number, then write the recommendations and asks yourself.
- After the meeting, have AI turn your notes into the follow-up email and the CRM actions, and review both before sending.
#When is a template and ChatGPT enough?
If one or two CSMs run a few dozen accounts and the data sits in one CRM, a good template, a CRM export and an approved AI tool will get you most of the way. You do not need a build. You do need a rule on which customer data may go into which tool, and the habit of checking figures.
A built workflow starts to make sense when many CSMs prep many reviews, the data sits in several systems, or the copy-paste step is where customer data leaks into personal tools. At that point the export and the draft should run inside your own accounts, with a log of what was produced.
#How Stennir runs this
Stennir starts with AI consulting when the team is unsure whether QBR prep is the right first workflow; the AI Opportunity Audit ranks it against renewal prep, onboarding and escalations. With AI implementation, Stennir builds the QBR draft into the CRM and tools you already use, in Arabic and English, hosted where your rules require, with the CSM approving every deck.
AI training through the AI Operator Program then teaches CSMs on their own accounts: how to check a draft, what to ask for, and where AI should not be trusted. Usage is measured before and after, and the program ends with an adoption report.
We are
Stennir builds QBR prep into your own CRM and accounts, keeps the CSM approving every deck, and hands you the off switch.
We aren't
Stennir is not a customer success platform you rent, and it does not send anything to your customers without a person's sign-off.
#What should a head of customer success do next?
Pick the next round of reviews and time the prep honestly, section by section. That tells you where AI would help. See how this fits a wider support and success team at AI for customer support, and how call notes become actions in AI for meeting notes. Then take the free readiness assessment or book a 30-minute call.
#Common questions
What do you need from us to build QBR prep?
Your QBR template, read access to the CRM, ticketing and usage data, two or three past decks you were happy with, and one CSM lead who can explain what a good review looks like. You also name who approves the draft. Stennir maps the rest with that lead before building anything.
Will the customer see an ai-written deck?
They see a deck the CSM wrote and checked. AI assembles the data and a first draft of the factual sections, but the narrative, risks and recommendations come from the CSM. Nothing reaches the customer until that person has reviewed every figure against its source and approved the final version.
Where does our customer data stay?
In your accounts. The workflow runs in your CRM and cloud, hosted in the region your contracts and data rules require, including in-country hosting where needed. You own the accounts, the data and the off switch. Personal data rules such as the UAE and Saudi PDPL laws are covered during mapping.
Does it work for arabic-speaking accounts?
Yes. Arabic and English are fully supported, so call notes, tickets and emails in either language feed the same draft, and the deck can be prepared in the language the customer prefers. The CSM still reviews the wording, especially names, titles and any figures quoted from Arabic documents.
