[Un]Churned · 23 Sep 2026 · From the week of 21 September
Notion's Outcomes Architects on Why Today's CSMs Are Builders
These are notes on the conversation, checked against its transcript. The episode itself has the full discussion.
In brief
Host Josh Schachter of Gainsight interviews three members of Notion's Americas Outcomes Architecture team, the group formerly called customer success: leader Christina Parra, mid-market outcomes architect Ariel Risman (about 30 accounts) and commercial outcomes architect Kellie Woodin (about 90 accounts). They describe how Notion merged its relationship/renewal managers into CS, added technical building to the role, and renamed CSMs 'outcomes architects' who co-build workflows with customers in workshops instead of running one-to-many trainings. They cover an AI maturity model embedded in 'AI transformation plans', proving ROI credit by credit under usage-based pricing, and what CS has learned from owning renewals. They also cover hiring for curiosity, agency and urgency, and internal agent systems such as a multi-agent feature-request pipeline. The core argument is that CS now sells work rather than software, and that anchoring co-built workflows to customer business objectives drives adoption, removes the need to re-justify value, and opens new expansion levers.
For founders
- Notion reorganized go-to-market from nine sales teams into roughly four functions: solutions consultants as technical builders on the land, outcomes architects post-sale, and forward deployed engineering for complex multi-agent work.
- Notion merged relationship and renewal managers into CS so the team owning adoption also owns commercials; the team says the value conversations about workflows shape how customers think about the partnership.
- With usage-based credit pricing, Notion's CS team says a seat contraction can be offset by building custom-agent workflows that raise credit consumption, so there is no ceiling on account value.
- Christina Parra now hires CSMs mainly for curiosity, agency and urgency rather than years of experience.
- Notion built an internal multi-agent pipeline that captures feature requests from Gong and Slack, checks Salesforce, and ranks the top 10 requests across customers.
For revenue leaders
- Replace one-to-many trainings with co-build workshops that leave customers with working workflows; Notion runs them for up to about 190 attendees virtually and smaller exec or builder groups in person.
- Have sales document business objectives, stakeholders and one key workflow at handoff, then turn success plans into 'AI transformation plans' that diagnose maturity and set a mutual action plan.
- Prove ROI under consumption pricing with a simple calculator: hours saved times hourly cost versus the credits an agent consumes (Notion cites an email agent running on under 500 credits a day).
- In high-volume books, prioritize by who is renewing this quarter or next, who is onboarding, and who is mid-implementation, and push everyone else to async and self-serve resources.
- Anchor workflows to the customer's stated business objective so value doesn't need re-justifying at renewal.
What was said 22, most useful first
Notion consolidated nine sales teams into roughly four functions, replacing solutions engineers with technical-builder solutions consultants and treating 'post-sales' as obsolete. Listen
The go-to-market evolution took nine sales teams down to effectively four functions. Solutions consultants act as technical builders on the land motion, and outcomes architects handle what used to be called post-sales. A forward deployed engineering team handles the most complex technical, multi-agent workflows. The speaker said that in an agentic world there really is no post-sales, only continually showing value and adoption.
“we took nine sales teams and brought them down to effectively four functions. So we no longer have solutions engineers.”
Notion's co-build workshops pull in teams that weren't using Notion and feed the expansion motion. Listen
After the initial workflow is built, Ariel Risman uses workshops to increase cross-functional collaboration inside the customer. When workflows involving custom agents span functions such as product and revenue, other teams become interested, and outcomes architects then build workflows for teams that may not have been on Notion. He said the workshops also spark users to tinker and build on their own.
“we're actually building workflows for those respective teams that may not have been using Notion today. And that kind of feeds into this expand motion that we have as a business.”
Listen to the episode Retention & customer success Link to this
Consumption pricing puts 'good pressure' on CS to prove value, and Notion answers it with a calculator setting credit cost against hours saved. Listen
Kellie Woodin said product usage data doesn't tell the full story, so CS has to understand the actual time saved. Notion's internal 'custom agent hall of fame' includes a small calculator for what credits are worth. Her example takes hours an agent saves (about 10 a week for her email agent) times an hourly salary rate and compares that with credits that cost under $5 a day. Customers come to Notion for guidance on these calculations across their LLM and Notion credit usage, which she said positions CS as strategic.
“we've been able to include a small calculator of how to understand what the credits are worth.”
The value customers need proven is sometimes just reduced friction between teams, not hours saved. Listen
Ariel Risman said CS at Notion acts as a thought leader in helping customers define what value looks like, and it depends on the team's biggest issue at the moment. For some it is cutting time spent on administrative tasks. For others it is reducing friction between teams that didn't work well together, so they can move with the velocity AI demands. He said to identify those issues, improve them, then iterate.
“value could be just reducing friction between teams”
Listen to the episode Retention & customer success Link to this
Usage-based pricing gives CS new levers at renewal, such as offsetting seat contraction with agent-driven credit usage. Listen
Kellie Woodin said that under seat-based pricing, the best CS could do after a seat contraction was try to win seats back. Under Notion's new usage-based pricing there is 'no ceiling'. A contraction can be offset by building workflows with custom agents that raise credit usage. Her lesson from her first renewals was that they are no longer black and white: she brings in power users and other stakeholders to show what Notion can be beyond a license count.
“previously, if they had some kind of contraction in seats, the best we could do is try to get them back up.”
Notion runs a multi-agent feature-request pipeline that captures requests from Gong and Slack, checks Salesforce, and maintains a ranked top-10 list. Listen
A former solutions engineer designed the system, and CSMs use it constantly. An intake agent listens to Gong calls and monitors a Slack channel where CSMs post requests. Another agent restates the request in natural language, another checks Salesforce for whether the requester is a customer, and another logs it in a database. That gives a top-10 feature request list across customers, and requests are referenced in customers' AI transformation plans. The next step Christina Parra described as planned is an automatic alert to the customer when a request ships, to build trust.
“There's another one that checks Salesforce to see, is this a customer, is this not a customer?”
Notion's CS team reframed its job from selling software to selling work, building end-to-end workflows with customers instead of setting up use cases. Listen
Kellie Woodin said Notion's role shifted from developing use cases to developing workflows, because Notion is a blank canvas once tools, AI and agents are connected. The team now runs workshop-style co-build sessions rather than one-to-many trainings, so customers get end-to-end workflows running and start saving time right away. That is why the title changed from customer success manager to 'outcomes architect'.
“Now we're not selling software, we're selling work.”
Listen to the episode Retention & customer success Link to this
Notion merged its relationship/renewal manager motion into CS, and is now folding in technical building as well. Listen
Christina Parra described a first 'CS Evolution' that merged relationship managers and renewal managers with CSMs to bring the commercial piece into CS. The next evolution adds the technical piece. Her reasoning is that adoption, consumption, commercials and value are inextricably tied, so one team can be a consultative partner across all of them.
“we launched CS Evolution where we actually merged our relationship manager renewal manager motion with our CSM motion. So we brought together the commercial piece. And then this next evolution is bringing the technical piece in.”
Listen to the episode Retention & customer success Link to this
Agents now handle internal prep like QBR decks, and co-building produces immediate outcomes where training relies on users remembering steps. Listen
Agents now handle internal CS work such as building QBR decks, which frees time for customers. Traditional training ('here's how to log in') depends on people remembering what they were shown. Building alongside customers produces an outcome right away; her example was an executive report that took eight hours a week and is now automated.
“they've got a new executive report that used to take eight hours a week to put together and now it's automated.”
Listen to the episode Retention & customer success Link to this
Notion's sales team documents business objectives, stakeholders and one key workflow at handoff, so CS can carry the same language into its plans. Listen
Sales identifies the initial workflow during the land, and outcomes architects grow additional workflows as more teams come on. Notion focused on a consistent language between sales and CS: sales documents the business objectives, the stakeholders and one key workflow the customer wants to implement. That carries into the CS success plan, now called the AI transformation plan. Notion also uses MEDDPICC.
“making sure that our sales team is documenting, you know, the business objectives, the stakeholders and one key workflow that that customer wants to implement.”
Ariel Risman runs large virtual foundational AI workshops of about 190 people and smaller in-person sessions for executives and internal enablers. Listen
Ariel Risman described a foundational AI workshop on Zoom with about 190 attendees from revenue, HR, engineering, product and design, ranging from execs to ICs. In-person sessions are more exec-focused or aimed at people who enable others internally, such as technical program managers, IT or revenue enablement. These sessions sometimes involve drawing workflow diagrams on paper and then rebuilding them in Notion as Mermaid diagrams. Attendees leave with concrete workflows they can use.
“I had a foundational AI workshop with a team and there was about 190 people on a Zoom meeting.”
Listen to the episode Retention & customer success Link to this
Notion uses a four-stage AI maturity model: AI as thought partner, then personal assistant, then teammate, then the system itself. Listen
Customers are spread across all four stages, and each stage builds on the one before. Notion maps its workshops to the stages. Personal-agent sessions cover meeting notes and agents that read email and create task lists. Smaller cross-team sessions build multiplayer custom agents. Forward deployed engineering builds multi-agent operating systems for customers. The speaker said this also clarifies where CS builds beside the customer and where FDE builds for them.
“And then the real transformation is when AI becomes the system.”
Notion turned success plans into 'AI transformation plans' that diagnose maturity, set a mutual action plan, and prescribe workflows from a library of best-in-class customer builds. Listen
The plans keep the usual success-plan elements: business objectives, pain points and success metrics. They add a walkthrough of the AI maturity model to diagnose where the customer is and where they want to go. That feeds a mutual action plan listing the workflows to build together, plus recommended workflows based on the customer's objectives. The speaker said the growing library of best-in-class customer workflows lets the team be prescriptive, which customers are asking for. Some customers still get them labelled as success plans.
“We diagnose where they are and we have a conversation about like where do you want to go and how can we help you get there. And so then we pull together a mutual action plan.”
Listen to the episode Retention & customer success Link to this
Kellie Woodin manages a 90-account commercial book by sorting it into renewing, onboarding and mid-implementation buckets and giving the rest async and self-serve resources. Listen
Kellie Woodin said Notion moved from a 'scaled' segment to 'commercial', which she described as less scaled and more hyper-prioritized. Because the team now owns renewals, she buckets accounts by who is renewing this quarter or next, who is onboarding, and who is in an active workflow implementation, and gives those high-touch attention. The team relies on agents and operational rigor for velocity and leans on async and self-serve resources for everyone else. Workshops, onsites and office visits are reserved for selected accounts.
“who's renewing, who's onboarding and who we're in any workflow implementations with, and those become our prioritization buckets.”
Listen to the episode Retention & customer success Link to this
Kellie Woodin's email-drafting agent saves her roughly 10 hours a week and runs on under 500 Notion credits a day. Listen
Her agent, 'reply wood', drafts responses to customer product questions and best-practice requests overnight. She starts the day with about 20 drafts to edit instead of 20 unread emails. She estimated it saves an hour or more a day, 'maybe 10 a week', and said it runs on under 500 credits a day. A separate AITP agent drafts tailored AI transformation plans; these are less in-depth for commercial accounts than for higher-ARR mid-market ones.
“instead of having 20 unread emails, I have 20 drafts waiting for me to tweak or to add my flavor.”
A book spread across many industries lets her give customers proof points from comparable companies, and Ariel Risman adds that customers often learn from adjacent verticals. Listen
Kellie Woodin said Notion is not a point solution, so her book spans many industries, setups and workflows. Because she has so many customers, she can usually point a customer to a similar industry, size or setup she has worked with. Ariel Risman added that customers often want to learn from an adjacent vertical. His example was a facilities team learning from how a construction management firm uses agents to manage physical assets.
“they want to learn what maybe not someone in financial services is doing, but what's the construction management firm doing?”
Listen to the episode Retention & customer success Link to this
Translating a customer's top-level goal into specific workflows removes the need to re-justify value later. Listen
The speaker said workflows are anchored to the business outcome agreed from the start with the solutions partners. Their example: if a customer's goal for the next three quarters is releasing five new products, the outcomes architect sets up workflows that improve collaboration between engineering and GTM so they ship faster. Because the work is tied to that objective, value follows without a separate re-justification.
“There isn't like a re -justification because we're really tied and anchored to that business objective.”
Listen to the episode Retention & customer success Link to this
Christina Parra, who has worked as an AE, AM and CSM, says owning renewals strengthened her conviction that CS can and should own commercials. Listen
Christina Parra said owning renewals shows how important value conversations and human check-ins are. The value conversations about workflows translate into how the customer thinks about the partnership, which is why CS should own commercials. She said she always knew this from one side or the other, but feels it much more owning both sides.
“I think a lot of conviction that the CS world can and should own commercials”
Listen to the episode Retention & customer success Link to this
Christina Parra now hires CSMs mainly for curiosity, agency and urgency rather than years of experience. Listen
Christina Parra has hired CSMs for seven years across significant market changes. She said what stands out in successful CSMs now is curiosity, agency and urgency, plus willingness to learn and try things themselves. Her example is team members who, whenever Notion launches something, immediately work out how to use it in their own workflow. She looks for these qualities in addition to skills and calls them 'the game changer'.
“what stands out about CSMs that are successful now is curiosity, agency, urgency”
Christina Parra uses an internal 'Data Scout' agent for natural-language data analysis and escalates to data scientists only after trying it herself. Listen
Christina Parra described Data Scout as like having a data analyst sitting next to you who answers natural-language questions. As a CS leader she uses it to uncover insights about her business quickly. When she does need a data scientist, she arrives having already tried and run the data, which she says builds trust. She said multiplayer deep-agent use cases like this are where Notion will keep focusing, internally and with customers.
“imagine there's a data analyst sitting next to you.”
Notion is not turning CSMs into builders only; Christina Parra argues co-building still depends on core CS discovery skills. Listen
Asked whether the role is becoming purely technical, Christina Parra said 'definitely not'. Getting customer commitment to a multiplayer agentic workflow through a workflow diagram requires deep discovery: understanding the customer's business, tech stack, what they care about and what value they need to see. She called these core CS skills.
“You have to have deep discovery skills. You have to understand their business and their tech stack and what they care about”
Listen to the episode Retention & customer success Link to this
Many companies have 'experimental' AI adoption, and adoption gets stronger when workflows tie to business objectives where work already happens. Listen
Christina Parra said many customers have a C-level AI adoption initiative but struggle to tie it to the work actually happening. She said Notion's conversations succeed because work already lives in Notion and connects to Slack and Google. Workflows aligned to the customer's business objectives drive stronger adoption, whether an individual workflow or a multi-agent operating system.
“a lot of people have what I would call experimental AI adoption. This is much more pragmatic”