Key takeaways
- Define blocked as waiting on a decision, access, or artifact the team has not supplied.
- Name one first responder and one backup in the place the team already works.
- Give each escalation step a clock so a silent message has a consequence.
- Write the request with the task, the exact need, the impact, and the deadline.
- Review a month of escalations and remove the cause that repeats most often.
Define what counts as blocked
A blocked developer is waiting on a decision, access, or artifact that the team has not provided. The condition is different from a hard problem that needs quiet thinking time. Write the distinction down so a developer can report a block without sounding like they are complaining. A clear definition also stops the opposite problem, where a developer spends a day working around a missing decision instead of asking for it. The team should agree on the wording before the first sprint, not after the first missed date.
Common blocks include a missing repository permission, an unanswered contract question, a test environment that will not start, and a product decision that changes scope. Each one has an owner somewhere in the buyer organization, even when that owner is not obvious to the developer. Naming the categories makes the escalation path concrete rather than a general request for help. The developer should be able to point at a category and say what they need. That habit turns a vague complaint into a routed request.
Name the first responder
The first responder is the person the developer messages before anyone else. Choose someone who is online during the overlap window and who can either answer or route the question. A team alias is a weak choice because it hides whether anyone has read the message. A named person with a named backup gives the developer a real destination. The backup matters during leave, holidays, and the week a responder changes role.
Publish the responder in the same place as the sprint board, not in a document nobody opens. Rotate the role weekly and record the handover in a short message that names the outgoing and incoming person. The responder does not need authority over every decision, but they must know who holds each one. Their job is to move the request to the right owner within the first hour. A responder who cannot route is only a second place to wait.
Set a time limit for each step
Give each step a clock. The developer reports a block, the responder acknowledges within one business hour, and the decision owner answers within one business day. These numbers are examples rather than universal rules, and the team should choose limits that match the work and the time zones. The point is that a silent message has a consequence instead of sitting until someone notices. Write the limits beside the responder rota so nobody has to remember them.
When a step passes its limit, the request moves to the next level automatically. The developer does not need to decide whether to escalate, which removes the social cost of pushing a senior colleague. The responder or a delivery lead can make the move and should say so in the thread. Record who receives the escalation and what they are expected to do. An escalation that lands with someone who cannot decide only adds another wait.
Write the escalation message
A useful escalation states the task, the exact need, the impact, and the deadline. Replace a message that says we are stuck with one that says the staging login for the payments service fails, we need a working account, and the release check is due Thursday. The reader should be able to act without a call. Include what the developer already tried so the reader does not repeat the same steps. A screenshot or a log line is often faster than another paragraph of explanation.
Keep the message in a channel the buyer monitors, and link the ticket so the context survives the conversation. Avoid splitting the history across direct messages and a tracker. One thread with one owner is easier to escalate than a discussion spread over three tools. The responder should be able to forward the message to the decision owner without rewriting it. If forwarding requires a summary, the original message was not clear enough.
Protect decisions across time zones
A block that appears at the end of the offshore shift needs a route that does not depend on the buyer being awake. Record the question, the realistic options, and the recommendation before the shift ends. The buyer can then approve an option or choose another at the start of their day. This pattern turns a hard dependency on live conversation into a written decision. It also gives the next shift something to act on rather than a thread of questions.
Not every decision can be delegated, and that is fine. What matters is that the developer knows which decisions they may make alone and which need approval. Publish a short list for each workstream and update it when scope or risk changes. A developer who knows the boundary stops asking for permission on routine choices. The same list tells the responder when a request is genuinely an escalation and when it is ordinary work.
Keep the path visible to both teams
Put the escalation path in the onboarding material, the sprint board header, and the first-week checklist. A process that lives only in a wiki page is replaced by private messages within a month. The buyer contact and the provider lead should both be able to describe the first two steps from memory. Test that understanding with a new developer during onboarding. If they cannot say where to go, the document has not done its job.
Visibility also means the state of an open escalation is easy to see. Use one status field with clear values such as reported, routed, waiting on buyer, and resolved. Avoid a long comment history as the only record of where the request stands. A short status makes the weekly review possible without reading every thread. It also shows the buyer how many requests are waiting on their side of the boundary.
Review escalations for patterns
Collect the escalations for a month and look for repeats. Three blocked releases caused by the same access request point to a provisioning problem rather than three unlucky days. The review should produce one change to the process, the tooling, or the ownership map. Without the review, the same block returns with a new ticket number and a fresh round of messages. A small fix to provisioning can remove an entire category of delay.
Share the pattern with the buyer and the provider lead, not only the developers. The buyer controls access, priority, and product decisions, so most durable fixes need their participation. Measure the time from report to resolution and watch whether it falls after the change. A path that is written but never measured tends to drift back into improvised messages. The review is what keeps the written path honest.
Questions about assessing Philippine developers
How fast should an offshore developer escalate?
Report the block as soon as the developer can state the exact need and impact. Waiting a full day to look independent usually costs more than the message.
Who should be the first responder?
A named person in the overlap window who can answer or route, with a named backup. Avoid an unowned team alias.
What if the buyer is offline for a long stretch?
Write the question, options, and recommendation before the shift ends so the buyer can decide at the start of their day.
