draft-outreach

# Draft Outreach Generic outreach gets ignored. Personalized messages that reference real context — a recent interaction, a known pain point, or a relevant product already owned — get responses. This skill eliminates the blank-page problem by pulling Dataverse data to craft contextually relevant outreach, saving reps time while improving reply rates. It works within Dataverse without requiring external enrichment tools. ## Instructions ### Step 1: Identify Target Contact and Outreach Trigger Accept input from the user: - **Target:** Contact name, lead name, or account name - **Outreach type:** New intro, re-engagement, follow-up, cross-sell, renewal, or event-based - **Goal:** Meeting request, proposal follow-up, demo invite, check-in, etc. - **Tone:** Formal, conversational, or brief (default: conversational) #### Step 2: Retrieve Contact or Lead Record **If contact:** ``` SELECT contactid, fullname, firstname, jobtitle, accountid, emailaddress1, telephone1, description, createdon, modifiedon FROM contact WHERE fullname LIKE '%[name]%' ``` **If lead:** ``` SELECT leadid, fullname, firstname, jobtitle, companyname, emailaddress1, telephone1, description, leadsourcecode, leadqualitycode, createdon FROM lead WHERE statecode = 0 AND fullname LIKE '%[name]%' ``` Present matches if multiple found. Confirm the target with the user before proceeding. #### Step 3: Gather Account Context If a contact was found, retrieve their account: ``` SELECT accountid, name, industrycode, revenue, numberofemployees, description, address1_city, address1_stateorprovince, customertypecode, accountcategorycode FROM account WHERE accountid = '[accountid]' ``` #### Step 4: Review Relationship History **Recent activities (last 90 days):** ``` SELECT activityid, activitytypecode, subject, actualend, description, statecode, regardingobjectid FROM activitypointer WHERE regardingobjectid = '[contactid or accountid]' AND statecode = 1 ORDER BY actualend DESC ``` Limit to 10 most recent. Note: - Days since last interaction - Last activity type and topic - Any commitments made (from description) **Open opportunities:** ``` SELECT opportunityid, name, estimatedvalue, estimatedclosedate, salesstage, closeprobability, description, currentsituation, customerneed, customerpainpoints FROM opportunity WHERE customerid = '[accountid]' AND statecode = 0 ``` **Recently closed opportunities:** ``` SELECT opportunityid, name, estimatedvalue, actualclosedate, statecode, statuscode FROM opportunity WHERE customerid = '[accountid]' AND statecode IN (1, 2) AND actualclosedate >= '[90_days_ago]' ``` **Products owned (from won opportunities):** ``` SELECT opportunityproduct.productid, product.name, opportunityproduct.quantity, opportunityproduct.priceperunit, opportunity.actualclosedate FROM opportunityproduct JOIN opportunity ON opportunityproduct.opportunityid = opportunity.opportunityid JOIN product ON opportunityproduct.productid = product.productid WHERE opportunity.customerid = '[accountid]' AND opportunity.statecode = 1 ``` #### Step 5: Check Open Cases ``` SELECT incidentid, title, statecode, prioritycode, createdon, description FROM incident WHERE customerid = '[accountid]' AND statecode = 0 ORDER BY prioritycode ASC ``` Flag open cases — they affect tone (avoid aggressive outreach to accounts with open escalations). #### Step 6: Retrieve Open Tasks Linked to This Contact ``` SELECT taskid, subject, description, scheduledend, statecode FROM task WHERE regardingobjectid = '[contactid or accountid]' AND statecode = 0 ORDER BY scheduledend ASC ``` If any open tasks exist for this contact, confirm outreach aligns with existing follow-up plan. #### Step 7: Determine Personalization Signals Based on gathered context, identify the strongest personalization hooks ranked by specificity: **Tier 1 (most specific — use if available):** - Last call/meeting topic referenced by subject or description - Pain point mentioned in opportunity customerpainpoints - Commitment made in prior activity (action item) - Product recently purchased — now cross-sell opportunity **Tier 2 (account-level):** - Industry-specific context (manufacturing, financial services, healthcare) - Company growth signals (employee count, revenue) - Active opportunity stage and deal context **Tier 3 (generic fallback):** - Account category (prospect, customer, partner) - Time since last interaction - Lead source (inbound vs outbound) #### Step 8: Generate Outreach Draft **Select template based on outreach type:** --- **New Intro:** > Subject: [Company] + [Your Company] — [value proposition in 5 words] > > Hi [Firstname], > > I came across [Account Name] while [researching [industry] leaders / reviewing your profile] and noticed [specific detail from account description or industry]. > > We help [industry] companies [key value prop relevant to their industry]. [Reference a peer company result if available in Dataverse notes]. > > Would a 20-minute call make sense to explore if there's a fit? Happy to work around your schedule. > > Best, > [Rep Name] --- **Re-engagement (contact gone quiet):** > Subject: Re: [Last activity subject] — checking back in > > Hi [Firstname], > > It's been [n weeks] since we last connected about [last activity subject]. I wanted to check back in — [reference the context from last interaction]. > > [If open case exists:] I also noticed there's an open [issue] — I've looped in our support team to make sure it's resolved quickly. > > Are you still evaluating [solution/need from opportunity]? Happy to reconnect at your convenience. > > Best, > [Rep Name] --- **Follow-up (after meeting/call):** > Subject: Follow-up: [Meeting subject] — [key next step] > > Hi [Firstname], > > Thanks for your time [yesterday/on [date]]. A few things from our conversation: > > - [Key point 1 from activity description] > - [Key point 2] > - [Commitment / next step agreed] > > As discussed, [action item] — I'll have that to you by [date]. > > [If BANT gap identified:] Before our next conversation, it would help to understand [budget / decision process / timeline] so I can tailor our next steps appropriately. > > Best, > [Rep Name] --- **Cross-sell:** > Subject: [Product already owned] → [New product]: natural next step for [Company] > > Hi [Firstname], > > Given how you've been using [existing product] since [close date], I wanted to introduce [new product/solution]. > > Several [industry] customers who started with [existing product] have added [new product] to [specific benefit — e.g., reduce manual work in X process]. > > Worth a quick call to walk through what it could look like for [Account Name]? > > Best, > [Rep Name] --- **Tone adjustments:** - **Formal:** Replace contractions, add full titles, longer sentences - **Brief:** 3 sentences maximum, single CTA - **Conversational:** Default as shown above #### Step 9: Create Email Activity in Dataverse After generating the draft, offer to create an email activity record: ``` Use create_record tool with tablename: email - subject: [Generated subject line] - description: [Draft body text] - regardingobjectid: [Opportunity ID or Account ID] - directioncode: true (outgoing) - statecode: 0 (draft / open) - to: [Contact's activityparty reference] - from: [Current user systemuser reference] ``` Note: Email is saved as a draft (statecode = 0). Rep reviews and sends from Dynamics 365 or Outlook. #### Step 10: Provide Sending Guidance Along with the draft, provide: **Personalization highlights used:** - Which Tier 1/2/3 signals were incorporated and why **Suggested timing:** - If contact was active recently (<14 days): send today - If contact has gone quiet (>30 days): consider calling first - If open escalated case: wait for resolution before outreach **Follow-up cadence suggestion:** - If no reply in 3 days: follow up once - If no reply after 2 attempts: assign a task for manager review or route to nurture ### Output Format Provide: 1. **Subject line options** (2-3 variants from short to descriptive) 2. **Email draft** (full body, personalization signals highlighted) 3. **Sending notes** (timing, tone rationale, cautions) 4. **Draft saved confirmation** (Dataverse email record ID if created) ### Example Interaction **User Input:** "Draft a re-engagement email for Emily Torres at Fabrikam. We haven't talked in 18 days." **Skill Output:** ✉️ **Draft Outreach — Re-engagement | Emily Torres, Fabrikam** **Personalization used:** - Last interaction: "QBR discussion" — 18 days ago (Tier 1) - Open deal: Fabrikam Cloud Migration — $85,000, Propose stage (Tier 2) - Open case: API integration timeout errors (flag — acknowledge) --- **Subject options:** 1. Checking in: Cloud Migration timeline — Fabrikam 2. Quick follow-up from our last call, Emily 3. Fabrikam Cloud Migration — next steps? **Draft:** > Hi Emily, > > It's been a few weeks since we spoke about the cloud migration. I wanted to check back in — has anything changed with your timeline or the decision process on your end? > > I also wanted to acknowledge the API timeout issue that came in — our team is working to resolve it and I'll make sure you have an update by end of this week. > > If it helps, I can put together a revised proposal addressing the questions that came up in our last conversation. Would 20 minutes work later this week? > > Best, > Sarah --- **Sending notes:** - ⚠️ Open support case — acknowledge it early as shown; don't lead with the pitch - Confirm case resolution status with support before sending if possible - If no reply in 3 business days, follow up with a call (not another email) **Draft saved:** Email activity created in Dataverse (ID: `email-xxx`) — linked to Fabrikam Cloud Migration opportunity. Ready to review and send. ### Dataverse Tables Used | Table | Purpose | |-------|---------| | `contact` | Target contact details | | `lead` | Target lead details | | `account` | Account context and firmographics | | `activitypointer` | Relationship history | | `opportunity` | Open and closed deal context | | `opportunityproduct` | Products owned for cross-sell | | `product` | Product names | | `incident` | Open support cases | | `task` | Open follow-up commitments | | `email` | Save outreach draft as email activity | ### Key Fields Reference **email:** - `subject` (NVARCHAR 200) - Email subject line - `description` (MULTILINE TEXT) - Email body - `directioncode` (BIT) - Outgoing(1), Incoming(0) - `statecode` (STATE) - Draft/Open(0), Completed/Sent(1), Canceled(2) - `regardingobjectid` (LOOKUP) - Linked opportunity, account, or contact - `scheduledstart` (DATETIME) - Planned send time **contact:** - `firstname` (NVARCHAR) - For personalized salutation - `jobtitle` (NVARCHAR) - For context and tone calibration - `emailaddress1` (EMAIL) - Primary email - `description` (MULTILINE TEXT) - Notes and background ### Configurable Parameters - Default tone (formal, conversational, brief) - Re-engagement threshold days (default: 14 days without activity) - Cross-sell trigger (days since last purchase, default: 30 days) - Follow-up cadence (default: 3 business days) - Case severity threshold that pauses outreach (default: Priority 1 escalations) ## Examples ### Example 1: New Prospect Introduction **User says:** "Draft an intro email to Sarah Chen at Northwind" **Actions:** 1. Query contact record for Sarah Chen 2. Retrieve Northwind account context (industry, size) 3. Check for any existing activities or opportunities 4. Generate personalized intro email **Result:** ``` To: sarah.chen@northwind.com Subject: Quick thought on Northwind's supply chain initiative Hi Sarah, I noticed Northwind recently expanded into the Midwest market — congratulations on the growth. We've helped similar distribution companies reduce order processing time by 40% while scaling operations. Would you have 15 minutes this week to explore if there's a fit? Best, Mike --- Sending notes: • Personalized with company expansion context • Follow up in 3 days if no reply ``` ### Example 2: Re-engagement Email **User says:** "Help me reach out to James at Fabrikam again" **Actions:** 1. Query contact and account records 2. Review last interaction date and context 3. Check for open opportunities or past purchases 4. Generate re-engagement email referencing history **Result:** ``` To: james.wilson@fabrikam.com Subject: Following up on our cloud migration conversation Hi James, It's been a few weeks since we discussed your Q2 cloud migration goals. I wanted to check in — has the internal budget review wrapped up? I have a case study from a similar Financial Services company that might be relevant to share with your CFO. Happy to reconnect whenever timing works. Best, Sarah --- Context used: • Last contact: 18 days ago (demo call) • Open opp: Cloud Migration ($85K) • Stage: Proposal sent, awaiting budget approval ``` ### Example 3: Cross-sell Outreach **User says:** "Compose a cross-sell email to Alpine Ski House" **Actions:** 1. Query account and identify primary contact 2. Review products already owned 3. Identify complementary products 4. Generate cross-sell email with value prop **Result:** ``` To: lisa.park@alpineskihouse.com Subject: Maximize your CRM investment with Analytics Hi Lisa, Congratulations on 6 months with our CRM platform — your team's adoption rate is impressive. Many customers at your stage add our Analytics module to get deeper visibility into sales performance. Based on your 200+ open opportunities, you'd likely see immediate value. Want me to set up a 20-minute demo? Best, Mike --- Cross-sell context: • Purchased: CRM Platform ($42K) 6 months ago • Suggested: Analytics Module ($15K) • Open support cases: 0 (safe to approach) ``` ## Troubleshooting ### Error: Contact not found **Cause:** Name doesn't match any contact record, or contact is inactive **Solution:** - Search with partial name or company - Check lead table if contact not found - Present multiple matches for user selection ### Error: No relationship history available **Cause:** New contact with no activities or opportunities **Solution:** - Generate cold outreach template - Use account-level context if available - Note that intro is cold (no prior relationship) ### Error: Open support case blocks outreach **Cause:** Contact has active Priority 1 escalation **Solution:** - Alert user about open case - Suggest acknowledging case in email - Recommend waiting for resolution before sales outreach