Milin Desai

A launch or campaign can create attention, but repeat use depends on fixing hard bugs, supporting developers across time zones, and not breaking things while customers are asleep. The reasoning is operational: trust accumulates through the unglamorous work that keeps software dependable.

Desai shares a map showing one week of new paid organizations on Sentry and says the image still gives pause.

• The post's lesson is that the map is "earned," not a given: "You don't win it with a launch or a campaign.
• You win it by showing up, fixing the hard bugs, and not breaking things when people are asleep in other time zones."

Read original

The practical test is whether a developer can ground AI work in real telemetry—errors, traces, and logs—and trace a bad AI response back to user behavior. Desai also argues that AI observability should not be a separate silo; it should connect the AI feature to the wider application and feed issues into evaluation systems.

In one post, Desai recommends using Claude, Codex, or similar tools with Sentry telemetry through MCP or CLI as a "deterministic starting point" for fixes. In another, Desai reports that a Stack Overflow developer survey found most developers use Sentry for AI agent observability and security, then argues that capturing agent behavior is only the baseline; finding issues in traces and feeding them into evals is the harder work.

Read original · Original 2

Defaulting to an incumbent can look responsible because it saves a line item and avoids risk, but Desai argues that choice can hide long-term costs and leave the cleanup to the team that stays. The alternative is to judge the tool built for the job rather than the bundled option.

Desai describes a multi-six-figure deal won against a public incumbent that offered a competing product for free.

• The post's summary is "Free lost.
• Best of breed won."
• Desai says the buying team was playing offense and optimized for better business outcomes; the warning is aimed at seasoned executives who make short-term tooling decisions and move on before the bill comes due.

Read original

The reasoning is reciprocal rather than promotional: the maintainers do work that supports the product, so supporting them is a material priority.

Desai announces that the company gave $1,000,000 to open-source maintainers and thanks every maintainer. The post states the reason simply—support the builders of open-source software Sentry uses and promote the open-builder community—and links to details about where the money went.

Read original

A newcomer should start here because it shows the commercial argument underneath the product and community work.

Milin Desai describes winning a multi-six-figure deal against a public incumbent that offered a competing product for free, then draws the lesson: 'Free lost.

• Best of breed won.' The post says the buying team was playing offense and optimized for better business outcomes rather than the safe bundled option.
• Milin Desai also warns about seasoned executives who default to the incumbent, move to another role, and leave the cleanup to the team that stayed.

For brands planning a campaign

A developer-tools brand could use this as the basis for an executive interview or written point of view on build-versus-buy, tool consolidation, and long-term cost. The useful angle is a decision framework for teams choosing between a free incumbent and a best-of-breed tool, not a generic 'AI savings' message.

It is a useful early read because it states the operating belief behind the Sentry story.

Milin Desai shares a map of one week of new paid organizations on Sentry and says the image still stops them.

• The post argues that building a product developers trust is a slow game: 'You don't win it with a launch or a campaign.
• You win it by showing up, fixing the hard bugs, and not breaking things when people are asleep in other time zones.' The emphasis is on earned trust across global developer teams.

For brands planning a campaign

For a brand partnership, this points toward reliability, engineering quality, and customer-trust narratives rather than launch hype. A campaign could feature hard bug fixes, incident response, or behind-the-scenes reliability work, with the product benefit shown through the team's operating discipline.

A newcomer should read it to see the current AI positioning and the specific developer problem being addressed.

Milin Desai reports that in Stack Overflow's Developer Survey, most developers said they use Sentry for AI agent observability and security.

• The post argues that AI observability should not be a separate silo: teams want the whole user journey, not just LLM calls, so a bad AI response can be traced back to what the user did.
• It also says capturing agent behavior is now the baseline; the harder work is finding issues inside traces and feeding them into evaluation systems.

For brands planning a campaign

A B2B developer-tools brand could build a technical campaign around AI agent evaluation, trace-level debugging, and end-to-end observability. The post gives a concrete position to react to: observability should connect AI behavior to the wider application and user journey, not just monitor model calls.

Start here to understand the community-facing side of the work, separate from product positioning.

Milin Desai announces that the company gave $1,000,000 to open-source maintainers.

• The stated reason is simple: support the builders of open-source software the company uses and promote the community of open builders.
• The post thanks maintainers and links to details about where the money went.

For brands planning a campaign

This is evidence for community-led partnership ideas such as transparent open-source funding, maintainer stories, or co-created support programs. A brand wanting credibility with maintainers should focus on visible contribution and follow-through rather than a one-off sponsorship mention.

It is a good starting point for seeing the hands-on, tool-level vocabulary used in the work.

Milin Desai argues that developers do not need to overcomplicate AI outcomes.

• Amid talk of factories and loops, the post suggests using Claude, Codex, or similar tools with Sentry telemetry, including errors, traces, and logs, through MCP or CLI as a deterministic starting point.
• The point is to fix things in service of better outcomes for the business and users, with a light closing reference to 'Be like Scott Cooper.'

For brands planning a campaign

A campaign could support a practical tutorial or demo showing an AI coding workflow grounded in real telemetry: errors, traces, logs, MCP, and CLI. The value is a reproducible developer workflow, not an abstract AI productivity claim.

Start here to see the lighter, experience-led side of the work rather than the product or deal analysis.

Milin Desai connects a Sentry 'Hot Fix' video at Chase Center to a household love of basketball and the Golden State Warriors and Golden State Valkyries.

• The post says the video showed how Sentry helps improve the fan experience at Chase Center, then moves to Sentry serving 200,000+ software teams globally and to hiring builders with an ownership mindset.
• It ends with thanks to Nick Manning and Golden State and a note about rooting for the Valkyries in the playoffs.

For brands planning a campaign

For a brand trying to reach developer or sports-entertainment audiences, this supports experience-led content tied to a recognizable venue or team, with a clear human story and a recruiting call to action. It is not a straight product explainer, so the campaign should treat it as community marketing rather than technical proof.

A map showing one week of new paid organizations becomes the starting point for an argument that trust comes from showing up, fixing hard bugs, and not breaking things across time zones.

Starts from a single internal artifact—a map of new paid organizations—and turns it into a point about trust and retention.

For brands planning a campaign

A developer-tools brand could use this angle for executive commentary or customer-proof content about reliability, debugging, and support. It suits campaigns that need to explain why developers stay with a product after the initial launch.

Examples: Source 1

Short instructional post that names specific tools and describes a concrete developer workflow.

For brands planning a campaign

This supports practical tutorial or demo content showing AI coding assistants grounded in real production data. A developer-tools brand could use the angle to avoid abstract AI thought leadership and show a working workflow instead.

Examples: Source 1

The post argues that AI observability is not a separate silo: teams need to trace a bad response back to what the user did, treat agent tracing as a baseline, and feed issues into evaluation systems.

Pairs a survey claim with a product explanation and a set of priorities for developers building with agents.

The survey result is reported by Desai; the post does not include the underlying survey detail.

For brands planning a campaign

Useful for platform and engineering-leader content about end-to-end tracing, evaluation loops, and monitoring AI features inside a larger application. It gives a brand a way to discuss AI reliability without treating AI as an isolated product category.

Examples: Source 1

The argument is that the buyer chose the tool built for the job because it optimized for long-term outcomes, while an incumbent default can look responsible in the short term and leave the cleanup to the team that stays behind.

Uses a first-person sales anecdote, then generalizes it into a pattern observed over a long career.

The deal value and competitive outcome are Desai's account; the post does not independently verify them.

For brands planning a campaign

This is a strong fit for procurement or buyer-education content aimed at developer-tools founders and go-to-market teams. It could support narratives about total cost of ownership, consolidation risk, and positioning against free or bundled alternatives.

Examples: Source 1

Short announcement that states the total and the rationale, with a link to further detail.

The post gives the total and rationale, not the recipient list or selection criteria.

For brands planning a campaign

A brand could use this theme for community partnership or sponsorship content about maintainer support and open-source sustainability. It fits campaigns that want to reach developer communities through contribution rather than advertising alone.

Examples: Source 1

The post mixes an internal ritual with humor and a specific company detail.

Event post built around an internal ritual and a running joke about executive participation.

For brands planning a campaign

This supports employer-brand and behind-the-scenes content that humanizes engineering culture. A developer-tools company could adapt the format for a sponsored series or founder-led post about how its teams experiment.

Examples: Source 1

The post also cites Sentry serving more than 200,000 software teams and points to hiring.

First-person video story tied to a named sports property and venue.

The post describes the video and partnership; the video itself was not part of the reviewed text.

For brands planning a campaign

This is useful for sports or venue partnership case studies, experiential content, and employer-brand posts that cross over into consumer-facing storytelling. It shows how a B2B developer-tools company can make a local partnership concrete.

Examples: Source 1

The posts also carry a recruiting pitch pointing readers to Sentry's careers page. The material does not state Desai's job title or seniority.

View example

The relationship described is between Sentry and the team; the post does not say Desai has a personal commercial relationship with the Warriors.

View example

It reads as a personal and Sentry-customer reference, not a sponsorship or paid partnership.

View example

Stack Overflow is a data source in the post, not a business partner of Desai's.

View example

Knowlton argues for investing in an existing employee through a fractional leader rather than replacing them; Desai argues for choosing the tool built for the job rather than a free incumbent.

Knowlton focuses on people and leadership staffing for small businesses, while Desai writes about software selection, developer trust, and observability in developer tools.

Read original · Original 2 · Original 3

Strambach offers prompts and reporting queries for marketers; Desai recommends grounding AI work in existing telemetry instead of chasing new frameworks.

Strambach writes for marketers about paid media, attribution, and reporting, often in list or prompt formats. Desai writes for developers and engineering leaders about reliability, observability, and developer adoption.

Read original · Original 2 · Original 3

Both favor concrete choices over broad market claims.

Lemkin comments as an investor and interviewer across many companies and topics, from AI run rates to early product tactics. Desai speaks from inside a developer-tools company about developer trust, observability, and tool selection.

Read original · Original 2 · Original 3 · Original 4

Lemkin over Milin Desai when the campaign's audience is SaaS founders and executives deciding go-to-market strategy rather than engineering leaders judging tooling.

• In an October 2026 post recapping a 20VC conversation with Dev Ittycheria, Rory O'Driscoll and Harry Stebbings, Lemkin lays out the argument that developers show very little loyalty and switch fast while the enterprise is where a vendor sticks, and that a vendor missing from an AI-generated answer is effectively invisible.
• For a brief built around how AI agents now shape vendor selection, that framing sits closer to the buyer's question than product-level commentary.
• A second October 2026 post lists the early-stage practices that help a product — transparent pricing, free trials, POCs, SMB customers, one or two large customers, everyone doing support — which suits founder-facing content.

The tradeoff is first-hand operating detail.

• Lemkin's reviewed material is commentary on other people's companies and podcast conversations, not a report of running a developer-tools product, and it quotes guests rather than showing his own product decisions.
• It offers nothing comparable to Milin Desai's account of a multi-six-figure deal won against a free incumbent, the working definition of developer trust as fixing hard bugs and not breaking things across time zones, or the AI observability workflow that traces a bad model response back to what the user did.
• A campaign that needs an accountable practitioner speaking from inside a developer platform loses that grounding.

Examples: Source 1 · Source 2

His October 2026 posts are written for performance marketers: one gives a set of Claude prompts for connecting attribution data to revenue questions — which campaigns closed revenue, real ROAS counted on signed deals, what a qualified lead cost per channel, which journeys the deals that close have in common — and one is a 'plan versus reality' piece about budgets tripled after a competitor's ad, creative that never rotates, targeting that collapses to one job title, and LinkedIn reporting 47 leads while sales sees three emails. Those are concrete, checkable examples a demand-generation team could reuse in its own content.

Those examples do not answer the observability brief.

• Nothing in the reviewed posts covers developer trust, reliability, total cost of ownership against a bundled incumbent, or how engineering buyers evaluate tooling, so the campaign would need a different argument and a different audience.
• The formats are also built for fast consumption and lead capture — the prompt post ends by asking readers to comment for the full pack — rather than for the longer, evidence-led narrative an executive engagement requires.
• Strambach would win a brief aimed at marketers proving attribution; he would not carry a brief aimed at engineering leaders and CFOs choosing between free and paid infrastructure.

Examples: Source 1 · Source 2

His posts make that case directly.

• A December 2025 panel recap reports the line that fractional executives are 'insourced' rather than outsourced, that they take a seat at the table and report to the owner, and that the job is to work themselves out of a job without collapsing the company under growth.
• A March 2026 post argues that a loyal employee who has outgrown the role is better served by bringing in a fractional leader to coach them than by hiring a replacement.
• A company selling advisory retainers to owner-operated businesses would find the buyer psychology already articulated.

The subject matter does not transfer to a developer-tools campaign.

• There is no engineering, product, or developer-community content in the reviewed posts and no sign of the engineering-leader or CFO audience this brief targets.
• The campaign would have to supply the whole technical argument that Milin Desai's material already provides: the free-lost deal story, the reliability definition of developer trust, and the case that AI observability belongs inside the whole user journey.
• The audience is also smaller-business owners making a hiring decision, not software buyers comparing tools.

Examples: Source 1 · Source 2