International Sales Enablement

The opportunity

You want to connect what enablement does — sessions run, programmes shipped, people certified — to what sales does next: pipeline, win rate, ramp time. The honest split: AI is excellent at designing the measurement and explaining the data; it cannot reach into Salesforce and capture data on its own. Automated capture and a live dashboard are an IT and data-governance project. What you can do today is the part that usually stalls those projects: agree exactly what to measure and what the dashboard should show.

Try it in 15 minutes — a measurement model and dashboard spec

Run this in Copilot chat. You will get a metric tree linking activity to outcomes, plus a dashboard spec you can hand to whoever owns Salesforce reporting.

Act as an enablement measurement analyst. Help me design how we link International Sales Enablement activity to sales outcomes, and a dashboard spec I can hand to our Salesforce/reporting team.

Ask me up to 6 questions first about: the programmes we run, the sales outcomes leadership cares about (e.g. ramp time, win rate, pipeline coverage), what data we already have in Salesforce, and our reporting cadence.

Then produce:
1. A metric tree: leading enablement activity metrics -> intermediate behaviour change -> lagging sales outcomes, with the assumed link stated plainly for each.
2. A short list of the 5-7 metrics worth a dashboard, each with: definition, the Salesforce object/field it likely comes from, and how often to refresh.
3. A one-screen dashboard layout (what goes top-left, what to compare against).
4. The data we do NOT yet have and would need to capture, flagged separately.

Rules: do not invent metric values or claim a causal link is proven; label every activity-to-outcome connection as an assumption to validate. If a field is uncertain, say so rather than guessing the Salesforce schema.

Act as an enablement measurement analyst. Help me design how we link International Sales Enablement activity to sales outcomes, and a dashboard spec I can hand to our Salesforce/reporting team… (full prompt on the copy button above)

What we learned in the room

Salesforce exports can power a practical monthly review even before live dashboard automation exists. The pattern is: export the relevant opportunities, have Copilot apply agreed criteria (for example opportunity size, probability, owner, or stage), then draft a reviewed email to the right stakeholder using your approved template.

• Prompt first, agent second: refine the prompt interactively in Copilot until it reliably reads the export, applies the criteria, and drafts the email in the right tone. Then save that pattern as a recurring agent.
• SharePoint as the repeatable source: a monthly folder with the latest export and template gives the agent a stable place to look without asking people to re-explain the workflow each time.
• Manual send stays non-negotiable: Copilot can draft and personalise the emails, but someone still reviews and clicks send. That keeps judgement, accuracy, and stakeholder sensitivity in the loop.

Where an agent fits

Once the metrics are agreed, a Quarterly Enablement Impact agent can turn the same numbers into a narrative each quarter — “what we ran, what moved, what to do next” — grounded in your measurement model so it reports consistently. It writes the story; it does not invent the data.

The opportunity

Enablement is a constant flood of requests from many regions and stakeholders. AI is genuinely good here today — it can structure a stakeholder map, weigh competing requests against agreed criteria, and turn a messy inbox of asks into a ranked, defensible plan for the week.

Try it in 15 minutes — triage this week’s requests

Paste your real (or sample) list of open asks into Copilot chat.

You are my prioritisation partner for International Sales Enablement. I'll paste this week's open requests and in-flight work. Before ranking, confirm the criteria we should score against — propose a default set (impact on sales outcomes, reach/number of people, urgency/deadline, effort, strategic fit) and let me adjust the weights.

Then return:
1. A ranked table: request | who's asking | score by criterion | overall | recommended this week / next / defer / decline.
2. The 3 things to protect time for, and the 2 to push back on (with a one-line reason I can send).
3. Anything that looks duplicated or could be batched.

Do not invent deadlines or stakeholders I didn't give you. Where a request is too vague to score, list the one question you'd ask to score it.

[Paste your open requests below.]

Help me build a stakeholder map for International Sales Enablement (all international, excluding LAC). I'll describe the key stakeholders and regions. For each, capture: what they care about, their influence on our priorities, how they prefer to be communicated with, and the current state of the relationship. Then suggest a light cadence for keeping each warm. Flag any stakeholder where I've given too little to map, rather than filling it in. [Describe your stakeholders below.]

You are my prioritisation partner for International Sales Enablement. I’ll paste this week’s open requests and in-flight work… (full prompt on the copy button above)

What we learned in the room

Copilot can help turn a lightweight Excel tracker into a working prioritisation system: project name, category, status, owner, key dates, and links back to source material. From there, it can summarise what changed, generate a simple dashboard, and produce a daily or weekly priority read-out.

• Excel is a useful starting point because the team can own it immediately, then export or import the structure into Monday.com if that becomes the central record.
• Scheduled prompts are useful for recurring nudges — for example a weekly priority email or chat summary based on the current tracker.
• Centralisation matters: Notebooks, shared folders, and trackers need a team home, otherwise every good workflow becomes one person’s private system.

Where an agent fits

Save the triage pattern as a Weekly Prioritisation agent grounded in your scoring criteria and connected to your Microsoft 365 inbox and Teams, so every Monday you paste the week’s asks and get the same ranked plan in your format.

The opportunity

This is the single biggest time sink you named, and it is where AI pays back fastest. The ~3 weeks is mostly drafting structure, writing content, and polishing presenters’ slides — all first-draft work AI does well, leaving you the judgement: accuracy, sequencing, and the room. The win is not “AI builds the course”; it is “AI gets you to a strong draft on day one instead of day ten.”

Try it in 15 minutes — a 2-day course outline from a brief

Run in Copilot chat. Give it the topic and audience; get back a structured outline you can react to.

You are an instructional designer for sales training. Help me build the outline for a 2-day in-person sales enablement course. I'll give you the topic, the audience, and the outcomes. Ask me anything you need first (max 6 questions).

Then produce:
1. Learning outcomes (what people can do by the end), written as observable behaviours.
2. A 2-day agenda broken into sessions with timings, each tagged: teach / practise / discuss / assess.
3. For each session: the key points, a suggested activity, and what "good" looks like.
4. A pre-work list and a list of materials to prepare.

Keep it practical for adult learners — minimise lecture, maximise practice. Mark anywhere you're assuming something I didn't tell you, and don't invent MillerKnoll product facts or figures.

You are an instructional designer for sales training. Help me build the outline for a 2-day in-person sales enablement course… (full prompt on the copy button above)

Presenter PPT refinement

For presenters’ slides, do this in Copilot in PowerPoint (desktop), on the MillerKnoll template, so it stays on-brand. Open their deck and run prompts like these in the Copilot pane:

Review this deck against the slide layouts in this file. Flag slides that are text-heavy, off the MillerKnoll template, or unclear, and suggest a fix for each — using only layouts already in this file.
Rewrite the bullets on slides [x–y] to three crisp points each, and move the detail into speaker notes.
Add speaker notes for any slide that has none, in a confident, plain presenting voice.
Propose a tighter running order for these slides for a 90-minute session, and say what to cut if we're short on time.

Deeper, step-by-step version: the Copilot in PowerPoint course.

What we learned in the room

Copilot in PowerPoint works best when it is opened inside the deck or template you want it to use. That keeps its context narrower: it sees the current file, the slide layouts, and the existing structure, so it is better at staying on-template than a generic chat draft.

• Start from the right container: open the MillerKnoll template or a strong previous deck first, then ask Copilot to create, rewrite, or reformat inside that file.
• Localise before polishing: use Copilot to adapt US-centric copy, headers, examples, and tone for international audiences, then review for nuance and market fit.
• Use PowerPoint for layout and movement: Copilot can suggest slide structure, update text, and add simple animations or transitions. For heavier content generation, draft with a Copilot agent or Notebook first, then bring the refined content into PowerPoint.

Where an agent fits

A Course Builder agent grounded in your house training format and a couple of approved past courses will draft new outlines in your structure every time — so the three weeks becomes a few focused days of expert review.

The opportunity

You already gauge knowledge before and after with Mentimeter quizzes (Charlotte can advise on how you run them today). AI helps on both ends: it can write a well-balanced quiz from your course content — matched pre/post so the comparison is fair — and it can analyse the results afterwards to show what shifted and what did not land. The quiz can still be delivered in Mentimeter (or Microsoft Forms); AI does the authoring and the read-out.

Try it in 15 minutes — a matched pre/post quiz

Paste your course outline or content into Copilot chat.

You are an assessment designer. From the training content I paste, write a knowledge check to run before and after the session. Produce:
1. 8-10 questions that test the actual learning outcomes (not trivia), a mix of multiple-choice and one or two short scenario questions.
2. A matched pre-version and post-version: same constructs, lightly reworded so people can't just memorise, so the before/after comparison is fair.
3. The correct answer and a one-line rationale for each.
4. A note on which outcome each question maps to.

Only test things that are actually in the content I gave you. If the content is thin on an outcome, tell me rather than inventing a question. [Paste your course content below.]

I'll paste the pre- and post-training quiz results (anonymised). Analyse the shift: which outcomes improved most, which barely moved, and where people are still weak after the training. Give me a short read-out I could share with the team and one or two changes to the course this suggests. Use only the numbers I provide — don't estimate beyond the data, and flag if the sample is too small to draw conclusions. [Paste results below.]

You are an assessment designer. From the training content I paste, write a knowledge check to run before and after the session… (full prompt on the copy button above)

The opportunity

Onboarding is the same content delivered to many people in different regions and roles — exactly the shape AI scales well. The move is: write the core programme once, then let AI personalise it per role/region/experience level and keep it current as products and process change. A grounded agent can also act as an always-on coach answering new joiners’ questions from your approved material.

Try it in 15 minutes — personalise an onboarding plan

Paste your standard onboarding plan into Copilot chat.

I'll paste our standard sales enablement onboarding plan. Personalise it for a specific new joiner: [role], [region], [experience level], [anything specific]. Keep all the must-do items, but re-sequence, re-weight, and adjust examples to fit this person — and flag anything in the standard plan that won't apply to them. Produce a week-by-week 30/60/90 view with clear "by now you can…" checkpoints. Use only what's in the plan I gave you; don't invent modules we don't have. [Paste the standard plan below.]

I’ll paste our standard sales enablement onboarding plan. Personalise it for a specific new joiner… (full prompt on the copy button above)

What we learned in the room

Notebooks are a useful bridge between messy feedback and reusable training assets. They can hold onboarding feedback, useful chats, source documents, and research in one place, then become the working context for curriculum guides, presentations, or updated enablement materials.

• Summarise feedback where it already lives: Copilot can search email and Teams conversations for onboarding feedback, pull out themes, and turn them into updates for a Notebook.
• Keep chat value from disappearing: when a Copilot chat produces useful research or wording, save the relevant output into a Notebook so the team can build on it later instead of starting from scratch.
• Use the right app: some Notebook features may be easier or only available in the Microsoft 365 app rather than Teams, so troubleshoot access there before assuming the feature is missing.

Where an agent fits

A Sales Onboarding Coach agent, grounded in your onboarding materials and playbooks, gives every new joiner the same high-quality, always-available guide — and you update one set of source documents instead of re-cutting decks. When the source changes, the coach changes with it.

Open Agent Builder on your computer (not on your phone — Agent Builder needs a desktop browser):

1. Go to microsoft365.com/chat.
2. In the left menu, click Agents, then New agent, then Skip to configure.
3. You will work on the Configure tab (fill in the fields), then Try it (test your agent), then click Create when you are ready to save.

The opportunity

This is the one path that is mostly platform configuration, not agent-building. Microsoft Teams can show live translated captions during a meeting, and Copilot can produce a translated recap and summary afterwards — covering a lot of what you currently outsource for international calls. It is worth a structured pilot against your real use cases before deciding what, if anything, still needs the external vendor.

What to pilot

• Live translated captions in Teams meetings — check which languages your team needs are supported, and whether this requires Teams Premium on your tenant.
• Teams interpreter — in the meeting controls, go to Record and transcribe → Language and speech, then turn on Interpreter and set the desired language. Language coverage has expanded noticeably in the last six months; confirm your tenant supports the languages you need.
• Attendee vs. organiser setup — test whether participants with Copilot licences can enable translation for themselves (not only the meeting organiser). This matters for regional colleagues who need French, Italian, Mandarin, or other languages on internal calls.
• Copilot meeting recap translated into the languages people need afterwards — you can also take the transcript and ask Copilot to translate it into another language.
• Quality and trust: run a real bilingual call and have a fluent speaker judge whether captions are good enough for the stakes of the conversation. Translation quality varies by language pair and by jargon.

What we learned in the room

The team confirmed Copilot licences are in place, so Teams interpreter and live captions are worth testing immediately on internal calls. The group clarified that both organisers and attendees with the right licence should be able to activate translation features, while external guests may face limits. A few constraints surfaced that shape the pilot:

• Dealer huddles are different from internal calls. Many dealer participants join as guests and may not have the Copilot licence that unlocks interpreter features — this is where you still use Interpretify today.
• Interpretify works but is operationally heavy — calls need to be set up well in advance, and the workflow feels clunky compared to turning on interpreter in a Teams meeting.
• MillerKnoll terminology is the main quality risk. Product names, internal acronyms, and domain terms (e.g. a word like “concert” that means something specific in your context) can trip up machine translation. With Interpretify you could submit a terminology list in advance; with Teams the practical next step is to build a glossary/acronym library, run translation on real calls, note where it stumbles, and feed those terms into Copilot when you ask it to translate a meeting recap.
• Voice simulation (interpreter speaking in a simulated voice of the presenter) is available but may not suit every use case — captions and translated recap may be the better starting point.

Next step: build the acronym/terminology library, test Teams interpreter on an internal call with a fluent speaker judging quality, then decide per use case — internal team calls, dealer huddles, high-stakes negotiations — where Teams is enough and where Interpretify or a human interpreter still earns its place.

Use AI to plan the pilot

Help me plan a small pilot to test whether Microsoft Teams live translated captions, Teams interpreter, and translated Copilot recaps could replace or reduce the external live-translation service (Interpretify) we use for international sales calls.

Produce:
1. Questions for IT about Teams/Copilot licensing, supported languages, and whether attendees (not just the organiser) can enable interpreter on their own.
2. A simple test plan: which internal calls vs. dealer huddles to test, which languages, and how a fluent speaker should judge quality.
3. A guest/dealer access checklist — what works when participants join as guests without a Copilot licence.
4. A glossary/acronym preparation step: how to collect MillerKnoll product names and internal terms, test where translation stumbles, and feed that list into Copilot recap translation.
5. Risks to watch: accuracy on jargon and product names, confidentiality, accessibility, and operational overhead vs. Interpretify.
6. A simple scorecard to decide go / partial / stay with vendor for each use case (internal team call, dealer huddle, high-stakes negotiation).

Don't assume our licensing or language support — list both as things to confirm with IT.

The opportunity

This is the most agent-forward theme, and the one you are most keen to explore. Three moves sit underneath it: generate and localise content for each market, recommend the right content to a seller or dealer for their situation, and let people self-serve answers instead of emailing enablement. An agent grounded in your approved content library does all three — and it is the natural home for a virtual version of the Sales Method card deck.

The Sales Method deck, as an agent

The physical card deck works because it is a structured way to apply the method to a real deal. An agent grounded in the deck’s content turns that into something interactive: a seller describes their opportunity, and the agent walks them through the relevant cards, asks the method’s questions, and helps them apply it — the deck you love, available on any call, in any language.

Try it in 15 minutes — scope the Sales Method agent

Run in Copilot chat to get a paste-ready agent spec you can take into Agent Builder.

Help me scope a Copilot agent that turns our physical "Sales Method" card deck into an interactive coach for sellers. I'll describe how the deck works and the cards in it. Then produce a paste-ready agent spec:
1. Agent description (one paragraph).
2. System prompt / instructions: how it should guide a seller through the relevant cards for their deal, ask the method's questions one at a time, and apply the method — without inventing cards or steps that aren't in the deck.
3. 4 conversation starters a seller would actually use.
4. The knowledge it needs (which documents to attach in Agent Builder).
5. 3 test prompts, including one deliberately tricky case.

Interview me first if you need to understand the deck — one question at a time. [Describe the Sales Method deck below.]

I'll paste a piece of approved sales content. Produce localised versions for these markets/languages: [list]. For each: translate, but also adapt examples, tone, and any references so it lands locally — not a word-for-word translation. Keep all factual claims identical to the source; flag anything that needs a local fact I don't have, and never invent a local statistic or reference. List what a local reviewer should check before it goes out. [Paste the content below.]

Help me scope a Copilot agent that turns our physical “Sales Method” card deck into an interactive coach for sellers… (full prompt on the copy button above)

What we learned in the room

Agents and Notebooks are the collaboration layer around the content work. Agents can be shared with other Copilot-licensed colleagues, but the creator remains the admin who manages access and updates. Sharing with non-licensed users is limited, so access planning matters before an agent becomes part of a team process.

• Notebooks preserve project context — documents, research, chat outputs, and source material can live together so content creation does not restart every time someone opens a new Copilot chat.
• Agents work best after the workflow is proven — build the process in chat, save the useful outputs into a Notebook, then turn the stable pattern into an agent.
• Translation still needs expert review — AI can create a strong first localisation pass, but content or language experts should fact-check translated sales enablement documents before they go to sellers or dealers.
• The Sales Method opportunity is bigger than a deck — the group discussed an interactive sales enablement platform that connects the sales methodology, training resources, prompts, and adoption tracking instead of leaving the method as a static asset.

Where agents fit

• Sales Method Coach — the virtual card deck described above.
• Content Concierge — sellers and dealers ask “what should I send a client doing X?” and get the right approved asset, in their language.
• Localisation assistant — first-pass localisation for each market, with a human reviewer in the loop.

For self-service across dealers, scope starts simple (a grounded agent over approved content) and the harder questions — who can access what, where it lives — become a follow-up project. Start with the prototype; let it prove the value.