# Drivepoint's Add-In Agent ### Meet Your AI-Powered Agent in the Drivepoint Add-In The Drivepoint Add-In now includes an AI-powered agent that works directly inside your Excel SmartModel™. Instead of navigating menus or waiting on the team, you chat with the Drivepoint Add-In Agent in plain English and it takes action in your SmartModel™ in real time. {% embed url="" %} #### Quick Start: Your First 5 Minutes Open your SmartModel in Excel and launch the Drivepoint Add-In. Click on **"Chat"** inside the Sidepanel menu. Try these three prompts in order: **Step 1 — Get oriented** ``` Tell me about this model ``` The agent reads your SmartModel and gives you a summary: company name, model version, date range, active channels, and key metrics. This confirms the agent is connected and reading your data correctly. **Step 2 — Pull in fresh data** ``` Import my latest Shopify sales ``` The agent fetches your most recent data from Drivepoint's data warehouse and writes it directly into your model. No exports, no copy-paste. You can also try: ``` Pull in my QuickBooks data ``` ``` Import retention predictions ``` **Step 3 — Ask a question** ``` What was revenue last month? ``` The agent reads from your model and responds in plain English. You're up and running. *** #### What You Can Do The agent can help you understand your model, pull in fresh data, answer questions about your numbers, and build the schedules and reports you need. Here's what you can do today. > **Tip:** You don't need to memorize exact phrasing. Ask the way you'd ask a teammate. If the agent isn't sure what you mean, it'll ask a clarifying question. *** #### Model Ops Understand, check, and maintain your SmartModel. **Summarize Model** Get oriented in any SmartModel — company, version, date range, channels, and key metrics. Try these: ``` Tell me about this model ``` ``` Summarize this model ``` ``` Give me an overview ``` ``` Stand up the model ``` **Interrogate Model** Trace any number back to its source. Select a cell first, then ask the agent to walk you through the logic. Try these: ``` Where does that number come from? ``` ``` Why is revenue $X? ``` ``` Walk me through this ``` ``` What's driving that line? ``` **Tip:** Click on a cell before asking "trace this" or "what's this?" — the agent uses your selection as context. **Keep Your Model Healthy** The agent handles the maintenance work so you don't have to — auditing, fixing broken references, optimizing, and navigating your workbook. **Audit before you send.** Sanity-check a model before it goes to a stakeholder, investor, or board. ``` Audit this model ``` ``` Validate before I send ``` **Fix broken references.** Scan for `#REF` errors, broken formulas, and formatting inconsistencies — then repair them. ``` Do I have any broken references? ``` ``` Fix the broken formulas ``` **Speed up slow workbooks.** Diagnose volatile formulas and reduce file size. ``` The model is slow ``` ``` Reduce file size ``` **Navigate the workbook.** Open any tab or find what you're looking for. ``` Open the P&L tab ``` ``` Where do I change CAC? ``` **Cell Edits & Formulas** Need to make a quick change or build a formula? Ask the agent. Try these: ``` Change cell B5 on the Settings tab to 12 ``` ``` Write a SUMIFS formula to total revenue for January ``` *** #### Data & Workflow Keep your model current with live data from Drivepoint's warehouse. **Import Data on Demand** Tell the agent to pull in your latest data and it writes directly into your model from Drivepoint's data warehouse. Try these: ``` Import my latest Shopify sales ``` ``` Pull in my QuickBooks data ``` ``` Import retention predictions ``` ``` Pull in my Amazon sales data ``` **Tip:** You don't need to know which import to use. Just describe what you need — "I need my latest sales data" — and the agent will recommend the right import for your setup. **Handling unmapped accounts:** If your financial accounts aren't fully mapped yet, the agent will let you know and redirect you to the mapping screen so you can resolve it before importing. **Imports with breakdowns:** Some imports require input from you — for example, how you want product data grouped. The agent will ask you before proceeding. **Roll Your Model Forward** One prompt and the agent moves your SmartModel to the next period, handling tab logic, references, and everything in between. Try this: ``` Roll forward to the next month ``` **Tip:** After rolling forward, follow up with a data import to pull in the latest actuals for the new period. The agent handles both steps from the chat. **Shortcut:** If your books are closed, try: ``` My books are closed ``` The agent will import the latest QuickBooks information and ask whether you'd like to roll the model forward — handling both steps in one flow. *** #### Analysis Turn your model into answers. Ask about any metric, time period, or line item in plain English. **Ask Questions About Your Numbers** The agent reads your model and answers in plain English. Try these: ``` What was gross margin last quarter? ``` ``` Which channel has the highest margin? ``` ``` How does my DTC revenue compare to Amazon this month? ``` ``` What's our total inventory? ``` **Tip:** Be specific with time periods and line items. "What was COGS in February?" gives you a sharper answer than "Tell me about costs." **Variance Analysis** Plan vs. actual, pacing, and pre-rolls. Try these: ``` Why did we miss? ``` ``` Budget vs. actual ``` ``` Are we on track? ``` ``` Give me a pre-roll ``` **Margin Analysis** Profitability by channel and mix. Build margin bridges and diagnose margin decline. Try these: ``` What are our margins? ``` ``` Amazon vs. DTC profitability ``` ``` Build a margin bridge ``` ``` Why are margins declining? ``` **Product Cost Analysis** Unit economics and COGS decomposition — landed cost per unit, bill of materials, cost structure. Try these: ``` Break down product costs ``` ``` What's the landed cost per unit? ``` ``` Walk me through unit economics ``` ``` Decompose COGS ``` **SKU Rationalization** Find what's working and what to cut across your product portfolio. Try these: ``` Run a SKU rationalization ``` ``` Which SKUs are dragging margin? ``` ``` Which products should we kill? ``` ``` Do we have too many SKUs? ``` **Cohort Analysis** Retention, LTV, CAC payback, and repeat purchase rates. Try these: ``` Show me cohort retention ``` ``` What's our LTV:CAC ratio? ``` ``` What's our payback period? ``` ``` What's our repeat purchase rate? ``` **Marketing Efficiency** CAC, ROAS by channel, and ad spend diagnostics. Try these: ``` What's our CAC? ``` ``` What's our ROAS by channel? ``` ``` Where should we cut marketing? ``` ``` How efficient is our ad spend? ``` **Inventory Analysis** Stock levels, weeks of supply, reorder signals, and dead stock. Try these: ``` Are we at risk of stocking out? ``` ``` What's our weeks of supply? ``` ``` What do we need to reorder? ``` ``` Do we have dead stock? ``` **Trade Spend Analysis** Promos, deductions, and retailer-level P\&L. Try these: ``` Analyze trade spend ``` ``` Are our promotions working? ``` ``` What's our deduction rate? ``` ``` Show me retailer profitability ``` *** #### Builder Create schedules, reports, and scenarios from a single prompt. **Build Schedule** Spin up revenue builds, COGS schedules, headcount plans, depreciation tables, and more. Try these: ``` Build a revenue build ``` ``` Make a COGS schedule ``` ``` Build a headcount plan ``` ``` Create a depreciation table ``` **Build Report** Board decks, monthly close reports, and investor updates. Try these: ``` Build a board deck ``` ``` Create a monthly close report ``` ``` Put together an investor update ``` ``` Package this up ``` **Create Scenario** What-ifs, stress tests, and runway levers. Try these: ``` What if we raise prices? ``` ``` What if CAC goes up 10%? ``` ``` What levers extend runway? ``` ``` Stress test this ``` **Compare Scenarios** Side-by-side plan comparison — budget vs. latest forecast, plan A vs. plan B. Try these: ``` Compare scenarios ``` ``` Which scenario is better? ``` ``` Budget vs. latest forecast ``` ``` Show me plan A vs. plan B ``` **Investor Readiness** Get your model raise-ready with due diligence checks. Try these: ``` Is this model investor-ready? ``` ``` Run a due diligence check ``` ``` What would an investor flag? ``` ``` Is this ready to share with VCs? ``` *** #### Tips for Great Answers * **Be natural.** Ask the way you'd ask a teammate — no special syntax needed. * **Point at something.** Select a cell or range first, then ask *"what's this?"* or *"trace this."* * **Iterate.** Follow up with *"go deeper,"* *"break that out by channel,"* or *"now compare to last quarter."* * **Ask for receipts.** If you want the logic, say *"show your work"* or *"explain the formula."* *** #### What's Coming Next We're actively iterating on the agent based on beta feedback. On the near-term roadmap: * **Streaming responses** — See the agent's answers render in real time instead of waiting for the full response. * **Bring your own skills** — Upload custom capabilities tailored to your team's workflows, conventions, and reporting templates. * **Your feedback, shipped** — The roadmap is being shaped by what we hear from beta customers. Tell us what's missing, what's clunky, or what would save you an hour a week. *** ### Need Help? Your Drivepoint rep can walk you through any of these workflows live. You also have a dedicated Slack channel for real-time feedback during the beta — drop questions, flag issues, or share what's working. We're building this with you. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.drivepoint.io/navigating-the-excel-add-in/drivepoints-add-in-agent.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.