Internal Comms & Meetings

How to prepare an all-hands deck as Small Marketing Teams

Internal Comms & MeetingsFor Small Marketing Teams3 apps11 steps~22 min to set up

Your three-person team is putting together an all-hands deck the same way every quarter: you pull pipeline numbers manually from HubSpot, screenshot the GA4 traffic chart, copy MQL counts into a slide, then ping the CEO for the company update copy that arrives 2 hours before the meeting. Someone rebuilds the campaign performance table from scratch in Google Slides because last quarter's version used different column headers. The whole thing takes 6-8 hours spread across two people, and you still walk into the all-hands knowing the HubSpot numbers are probably 3 days stale and the paid spend slide is missing LinkedIn Ads.

Build this on Starch → See the apps used

Internal Comms & MeetingsFor Small Marketing Teams3 apps11 steps~22 min to set up

Outcome

What you'll set up

A live marketing performance summary that pulls HubSpot pipeline data, GA4 sessions, and ad spend from Meta and Google Ads — queryable on demand so your slides reflect this week's numbers, not last Thursday's export

A Presentation Agent workflow (currently in beta — request access) that takes your data summary and a plain-English description and builds the slide deck structure, so you spend your time editing content instead of fighting layouts

A repeatable prep process: Starch surfaces the numbers, drafts the narrative, and hands you a deck that needs 30 minutes of review instead of a full day of assembly

The Starch recipe

Apps, data, and prompts

The combination of Starch apps, the data sources they pull from, and the prompts you use to drive them.

Apps used

Presentation Agent Growth Analyst Investor Reporting

Data sources & config

HubSpot is a scheduled-sync provider — Starch syncs your contacts, companies, deals, and owners on a schedule so pipeline data is always current. Gmail is also a scheduled-sync provider used for email thread context. Google Analytics 4, Meta Ads, Google Ads, and LinkedIn Ads are connected from Starch's integration catalog; the agent queries them live when your app runs. Notion connects from Starch's integration catalog for pulling campaign briefs and content calendar context. Presentation Agent is currently in development — request beta access to be notified at launch.

Prompts to copy

Pull this quarter's marketing metrics summary: total MQLs by channel, pipeline sourced by campaign, website sessions vs last quarter, and paid CAC broken down by Meta, Google Ads, and LinkedIn Ads. Flag any channel where MQL volume dropped more than 15% week-over-week and write a one-paragraph explanation of likely causes.

Build a 12-slide all-hands marketing update for a 120-person company. Slide 1: what marketing shipped this quarter. Slides 2-4: demand gen performance with pipeline contribution, MQL trends, and paid efficiency by channel. Slide 5: content and lifecycle highlights. Slide 6: what we're doing differently next quarter. Use the metrics summary I just generated. Keep the tone direct — no jargon, no vanity metrics.

Draft the 'what happened to MQL volume last month' section for our all-hands. Pull HubSpot lead data and compare to prior period. Reference any GA4 traffic shifts and note if paid spend changed. Write it as a 3-bullet executive summary suitable for a CEO to read in 90 seconds.

Run these in Starch → or paste them into your favorite agent

Walkthrough

Step-by-step

1 Connect HubSpot as a scheduled-sync provider. Starch syncs your deals, contacts, and owners on a schedule — pipeline contribution numbers in your deck will pull from this, not from a manual export you ran three days ago.

2 Connect Google Analytics 4, Meta Ads, Google Ads, and LinkedIn Ads from Starch's integration catalog. The agent queries each live when you run your marketing summary, so session counts and spend figures are current at the moment you ask.

3 Connect Notion from Starch's integration catalog if you keep your campaign briefs or content calendar there. Starch can pull the quarter's planned initiatives and compare them against what actually shipped.

4 Open the Growth Analyst app and run the metrics-pull prompt. Ask Starch to surface MQL volume by channel, pipeline sourced by campaign, paid CAC by platform, and any week-over-week drops above your threshold. This is the data layer your deck will sit on top of.

5 Review the flagged anomalies. If MQL volume dipped, Starch will surface likely causes from the data — traffic drop, spend cut, conversion rate change. Edit or confirm the explanation before it goes into the deck.

6 Draft the narrative sections. Ask Starch to write the 'what marketing shipped' section, the 'why MQL volume changed' explanation, and the 'what we're changing next quarter' bullets. Give it the tone constraint: direct, no vanity metrics, executive-readable in 90 seconds.

7 Use Presentation Agent (beta — request access) with the full prompt: describe the 12-slide structure, paste in your metrics summary, and let Starch build the initial layout with data visualizations and section headers. If Presentation Agent isn't available yet, Starch outputs a structured markdown outline you can drop into Google Slides in one pass.

8 Review slide 4 (paid efficiency by channel) against your actual LinkedIn Ads and Meta Ads numbers. Because Starch queried these live when you ran the summary, the figures should match what you'd see if you logged into each platform — but sanity-check the date range.

9 Send the HubSpot pipeline section to your head of sales for a 5-minute gut-check before the all-hands. Starch can generate a Slack message summarizing the key pipeline slides — paste it into your channel directly.

10 Lock the deck 24 hours before the meeting. If the CEO requests a last-minute metric update, re-run the Growth Analyst query rather than rebuilding the slide — Starch will pull fresh HubSpot and GA4 numbers and you update the one relevant slide.

11 After the all-hands, ask Starch to archive the marketing metrics snapshot to Notion — this gives you a historical record so next quarter you can show quarter-over-quarter trends without rebuilding the comparison manually.

See this running on Starch

Connect your tools, describe what you want, and the agent builds it. Closed beta is free.

Try it on Starch →

Worked example

Q1 2026 All-Hands Marketing Update — March prep

Sample numbers from a real run

MQLs this quarter (HubSpot)	312
Pipeline sourced by marketing ($)	840,000
Paid CAC — Meta Ads ($)	187
Paid CAC — Google Ads ($)	214
Paid CAC — LinkedIn Ads ($)	431
Website sessions (GA4, Q1 vs Q4 change %)	18
Email-sourced MQLs (Customer.io nurture)	74

It's March 27th. The all-hands is in two days. In past quarters, you'd spend today pulling a HubSpot pipeline report, screenshotting GA4, logging into Meta Ads and Google Ads separately to get spend and CPL, then manually calculating blended CAC in a spreadsheet before writing anything. This quarter: you run the Growth Analyst prompt at 9am. Starch queries HubSpot live and returns 312 MQLs for Q1, $840K in marketing-sourced pipeline, and flags that LinkedIn Ads CAC at $431 is running 2x Meta's $187 — with a note that LinkedIn spend increased 40% in February without a corresponding MQL lift. GA4 shows an 18% session increase quarter-over-quarter, but the email-sourced MQL number (74) is down from 91 last quarter, likely tied to the Customer.io sequence you paused in January. Starch drafts the 'why MQL volume changed' section as a 3-bullet executive summary. You edit two sentences. By 10:30am you have a complete data narrative. You hand the metrics summary to Presentation Agent with the 12-slide structure prompt and get a working deck by noon. You spend the afternoon on the 'what we're changing next quarter' slide — the only part that actually needs your judgment — instead of wrestling with table formatting.

Measurement

How you'll know it's working

MQL volume by channel (week-over-week and quarter-over-quarter)

Marketing-sourced pipeline ($) as a percentage of total pipeline

Paid CAC by platform (Meta, Google Ads, LinkedIn Ads) vs target

Website sessions vs prior period (GA4)

Hours spent preparing the all-hands deck (your internal benchmark)

Comparison

What this replaces

The other ways teams handle this today, and how the Starch version compares.

Google Slides + manual data pulls

Free and familiar, but every number is a manual export — you're always one stale screenshot away from presenting wrong data, and there's no connective layer between HubSpot, GA4, and your ad platforms.

Looker Studio

Good for always-on dashboards if you have time to configure it, but it doesn't write narrative copy or build slide decks — you still assemble the all-hands presentation by hand from the charts it produces.

Notion AI + manual data paste

Notion AI can clean up your writing, but it has no live connection to HubSpot or your ad platforms — you're still copying numbers in manually before it can help with the narrative.

ChatGPT or Claude with CSV exports

Works for one-off analysis if you export the right files, but there's no persistent connection to your live data sources — the process is manual every time and doesn't get faster with repetition.

On Starch RECOMMENDED

One platform — presentation agent, growth analyst, investor reporting all running on connected data. Setup in plain English; numbers stay current via scheduled syncs and live agent queries.

Try it on Starch →

FAQ

Frequently asked questions

Does Starch actually have a live connection to HubSpot, or do I have to export a CSV?

No CSV needed. HubSpot is a scheduled-sync provider — Starch connects directly to HubSpot and syncs your contacts, companies, deals, and owners on a schedule. When you run a query or build an app on top of it, the data is already in Starch's database and current as of the last sync.

What about GA4 and our ad platforms — Meta, Google Ads, LinkedIn Ads?

All three are available from Starch's integration catalog. You connect them once; the agent queries them live when your app runs. There's no scheduled sync for these — the data is pulled fresh at query time, which is fine for building a deck but means Starch isn't storing historical ad data over time.

The Presentation Agent is listed as 'currently in development' — what do I actually get today?

Today, Starch generates a complete structured outline of your deck with narrative copy, key data points, and slide-by-slide content — formatted so you can paste it into Google Slides in a single pass. Presentation Agent (the full visual deck builder with layouts and exports) is in beta; you can request access to be notified when it launches. Most teams find the structured outline cuts 80% of the prep time even without the visual builder.

Is Starch connected to Customer.io or Klaviyo for email performance data?

Both are available from Starch's integration catalog — connect either one and the agent queries it live when you need email-sourced MQL counts or campaign performance numbers for your deck. Starch doesn't store your ESP data long-term; it queries it at the moment you ask.

What if a data source I need isn't in the integration catalog?

Starch connects to 3,000+ apps through its integration catalog, plus any website through browser automation. If a specific reporting page or dashboard only exists in a web interface with no API, Starch can automate it through your browser — no API needed. That said, the tools a typical small marketing team uses (HubSpot, GA4, Meta Ads, Google Ads, LinkedIn Ads, Notion, Customer.io, Klaviyo) are all reachable today.

Is Starch SOC 2 certified? We'd be connecting HubSpot and ad platform credentials.

Starch is not SOC 2 Type II certified yet. That's an honest limit worth knowing before you connect production marketing data. If SOC 2 is a hard requirement for your company, that's a real constraint right now.

Can Starch pull the same metrics automatically every quarter so we're not re-prompting from scratch?

Yes — you can build a scheduled automation that runs your marketing metrics query on a set cadence (say, every Monday, or the last Friday of the quarter), formats the output, and posts it to Slack or saves it to Notion. You describe it in plain English and Starch builds the automation. Next all-hands, the data summary is already waiting.