Spotting Upsell Opportunities as a Developer
Applies to: every engineer on a customer engagement (contracting developers, forward deployed engineers, presales engineers) · Time to learn: about 25 minutes
You see the customer's real problems before anyone else does. When you observe one that the current scope does not cover, write it down with evidence and business impact, hand it to the engagement lead, and let the customer hear it as "an observation and an option", never as a price. Deliver the current scope fully first. Never hold back a fix, never start extra work for free, never commit dates or prices.
Why developers are well placed
- Trust. The customer's team works with you daily. They tell you what annoys them, what they are afraid of, and what they have postponed for a year.
- Visibility. You read the repositories, pipelines, dashboards, and cloud bills that no sales conversation ever reaches.
- Timing. You notice the problem while it is happening, with logs, tickets, and numbers to prove it.
The principle that keeps this honest: a good upsell is solving a real, observed customer problem. If you cannot name the problem, the evidence, and the business impact, there is no opportunity, only padding. The Agile Manifesto states the same spirit: "Customer collaboration over contract negotiation" and "Our highest priority is to satisfy the customer through early and continuous delivery of valuable software" (Agile Manifesto, principles).
The commercial side is covered in the sibling Sales Training. This page covers only the engineer's part: observing, validating, and handing over.
Signals to watch for
| Observed signal | Likely need | Where to read more |
|---|---|---|
| Deployments are manual, done by one person, or avoided on Fridays | Automated, repeatable delivery and a hardened runtime | Cloud-native development, How We Build: CI/CD |
| Infrastructure was clicked together in a console, no one knows the full picture | Infrastructure as Code and a documented landing zone | How We Build, Migration & Modernization |
| The cloud bill keeps rising and nobody can explain why | A structured look at cost, risk, and disposition of workloads | Cloud & Migration Assessment |
| Aging VMs, end-of-support platforms, or a monolith that blocks releases | Migration or incremental modernization | Migration & Modernization |
| Questions about where data is stored, who can access it, or which jurisdiction applies | Sovereign cloud controls | Sovereign Cloud on STACKIT |
| VMware licensing pressure or a wish for an on-premise cloud | Private cloud and hybrid connectivity | Hybrid Cloud & OpenStack |
| PoC meets its measured targets and users ask when they get it for real | Production hardening: gateway, evals in CI, tracing, SLOs | AI Productification, GenAI PoC |
| Staff repeatedly ask questions that the documentation already answers, or search is poor | Permission-aware retrieval over internal content | RAG Systems |
| Worry that prompts or documents must not leave the company | Private LLM platform on controlled infrastructure | Private & Sovereign AI |
| Several AI ideas, no agreed priority, unclear data readiness or governance | A short, scored assessment ending in a first use case | AI Assessment & Audits |
| A generic model misses domain vocabulary or format despite good retrieval and prompts | Check whether fine-tuning is the right lever | Fine-Tuning |
| A workflow needs multi-step tool use with approvals | Agent or workflow platform | Agentic AI Platforms |
| The customer wants AI assistants for their own developers but has no SSO, data controls, or policy | Rollout of AI tooling with governance | Cloud AI Platforms |
The table maps symptoms to candidate needs. It does not replace a conversation. Validate that the problem is real and matters to the customer before you write anything down.
Ethics: the ground rules
- Propose only what the customer needs. If the honest answer is "your current setup is fine", say so.
- Be transparent. The customer should always be able to see why you are raising something and what is in it for them.
- Never under-deliver the current scope to create follow-on work. The agreed deliverable gets your full quality, always.
- Never withhold a fix. If a defect or risk falls within the current scope, fix or report it now. If it falls outside, tell the customer plainly; do not sit on it until a later proposal.
- Keep the relationship first. A customer who trusts you will bring the next problem to you. That outcome is worth more than any single add-on.
How to raise it
- Finish and protect the current deliverable. An observation never displaces committed work.
- Observe and validate. Collect evidence: a log excerpt, a ticket count, a pipeline duration, a bill line. Check with a customer colleague that you understood the situation correctly.
- Write it down using the note template below.
- Raise it through the engagement's agreed channel, normally the engagement lead. They decide whether and how to take it to the customer.
- Propose to the customer as "an observation and an option". Describe what you saw and what it costs them. Leave pricing and dates to the people who own the commercial conversation.
Opportunity note template
# Opportunity note: <short title>
- **Date / author / engagement:**
- **Observation:** What did you see? One or two factual sentences.
- **Evidence:** Links, logs, tickets, numbers, with dates. Who at the customer confirmed it?
- **Business impact:** Cost, risk, time lost, or blocked goal, in the customer's terms.
- **Proposed option:** What could be done, and which service page applies.
- **Rough size range:** Small (days), medium (weeks), large (months). No prices.
- **Urgency:** Now / next milestone / when convenient, and why.
- **Relation to current scope:** Confirm that current deliverables are on track and unaffected.
Keep the note to one page. If you cannot fill in "Evidence" and "Business impact", keep observing before you raise it.
Timing
Raise opportunities at natural review points:
- Demos and sprint reviews, when the customer is looking at results and thinking about what comes next.
- Retrospectives, when friction is already on the table.
- PoC close, when the measured results support a decision. See PoC expectations and the case for MVP for how to frame that conversation; it is not repeated here.
- Milestone reviews and handovers, when scope, risks, and the road ahead are discussed anyway.
Hold it back during an incident, a release freeze, or a tense delivery week. The customer's attention belongs to the problem at hand.
Talking about it politely and concretely
| Pushy | Observation and option |
|---|---|
| "You really need a full migration." | "While working on the pipeline I saw that the three legacy VMs run an operating system whose vendor support ends this year. If useful, our team can outline options and effort." |
| "We can fix all of this for you." | "Deployments currently take about 40 minutes and need one person; I measured this over the last four releases (example figures). Automating them might be a sensible next step. Would you like us to write up an option?" |
| "Your current vendor did this wrong." | "The current setup works within its design. The constraint I noticed is X, and here are two ways teams commonly address it." |
| "Sign this and we start Monday." | "I will pass the observation to our engagement lead so you get a proper proposal with scope and timing." |
Pattern: what I saw, what it means for you, what the option is, who follows up. Use the customer's words and numbers, and ask before suggesting.
Scope protection and upsell
Boundaries and opportunities are two sides of one habit. When work falls outside the agreed scope, the change request process protects both parties. A change request the customer declines for now has not failed: the need is documented, the evidence is on file, and it can return as a proposal at the next milestone review. Likewise, never absorb out-of-scope work silently "to be nice"; it hides the real demand and erodes the next conversation.
What not to do
- Do not commit dates or prices. Those come from the people accountable for them.
- Do not start the extra work for free in the hope that it converts. It blurs the contract and sets the wrong precedent.
- Do not disparage the customer's current setup, their team, or other vendors. Describe constraints neutrally and respect the context in which earlier decisions were made.
- Do not go around the engagement lead. Parallel conversations confuse the customer.
- Do not trade on confidential information from one engagement in another.
Pitfalls to design around
- Lead with the problem. Start every conversation from the customer's pain and evidence, not from our service name.
- Keep the current deliverable green. Check status against the contract before any observation goes up the chain; see Accountability.
- Write before you speak. A one-page note forces clarity and gives the engagement lead something to act on.
- Size loosely, price never. Give a small/medium/large range internally and leave numbers to the owners.
- Use the agreed channel. Customer-facing proposals travel through the engagement lead; Customer communication covers tone and cadence.
- Accept a "no" gracefully. Record it, thank the customer, and revisit at the next review only if the evidence changes.
Checklist
- Current scope is on track and I am not trading quality for opportunity.
- I can state the observed problem in one sentence in the customer's terms.
- I have evidence (data, artefacts, a customer colleague's confirmation).
- The business impact is concrete and not inflated.
- The proposed option maps to an existing service page and genuinely fits.
- The opportunity note is complete and sent to the engagement lead.
- I have not promised dates, prices, or free work.
- Timing is a demo, retro, PoC close, or milestone review, not an incident.
- My wording is an observation and an option, with no pressure and no criticism of others.
Sources
- Agile Manifesto: value statement "Customer collaboration over contract negotiation" (fetched 2026-10).
- Principles behind the Agile Manifesto: "Our highest priority is to satisfy the customer through early and continuous delivery of valuable software" and "Business people and developers must work together daily throughout the project" (fetched 2026-10).
- Related internal pages: Contracts, scope and boundaries, PoC expectations and MVP, Customer communication, Accountability, and the linked service playbooks.