Excel sync was supposed to be a side feature. Customers made it the main one.

When we built Avery NXR, the Excel/workbook bidirectional sync was a "nice to have" feature. Useful but not central. We mentioned it on the website. We didn't lead with it.

Three months in, it's one of the most-used features by deployed customers. We've reordered some of our marketing because of how heavily it shows up in conversations.

Here's the story of why Excel sync mattered more than we anticipated.

What Excel sync actually does

Drop a .xlsx file into Avery NXR. The system:

→ Reads the sheets and converts them into Prisma models (database tables) → Reads formulas and converts them into TypeScript functions → Reads validation rules and converts them into Zod schemas

Now the data is queryable as a real database. Agents can read from it. Agents can write to it. When data changes in the database, the .xlsx file syncs. When the .xlsx file changes (someone edits in Excel), the database syncs. Bidirectional, with row-level conflict resolution.

It sounds like a technical curiosity. The reality is that this single feature unlocks something specific that customers value enormously.

What it unlocks

For most small/mid businesses, the "system of record" is a spreadsheet. Not a database. Not a SaaS platform. A spreadsheet.

→ The customer list is in a spreadsheet → The product catalog is in a spreadsheet → The pricing matrix is in a spreadsheet → The vendor list is in a spreadsheet → The hiring pipeline is in a spreadsheet → The project tracker is in a spreadsheet

For decades, this was fine. Excel is genuinely good at being a flexible system of record. The downside: spreadsheets don't talk to other systems easily. You can't really automate against them without exporting to a database first.

Excel sync changes the equation. Your spreadsheets BECOME a database, queryable by agents, while remaining usable as spreadsheets by the humans who own them.

What this looks like in practice

Customer A — a small consultancy. Their client tracker is a spreadsheet with 200+ rows: client name, status, project end date, billing cycle, point of contact, notes.

They built an Avery NXR agent that reads this sheet daily. The agent:

→ Identifies clients whose project end date is in <30 days (flag for renewal conversation) → Identifies clients who haven't been contacted in >14 days (flag for check-in) → Drafts personalized check-in emails for each → Emails the consultancy owner each morning with the action list

The spreadsheet stays the same. Owner still edits it directly when client details change. The agent reads from the same spreadsheet that's been the system of record for 5 years.

Total automation built in ~40 minutes. Owner doesn't have to migrate anything. Doesn't have to learn a new system.

Customer B — a 30-person product company. Their product pricing matrix is a spreadsheet with multiple tabs: base prices, regional pricing, volume discounts, custom enterprise quotes.

They built an Avery NXR agent that processes inbound RFPs and produces draft quotes. The agent:

→ Reads the RFP (PDF or email) → Identifies product, region, volume, special requirements → Queries the pricing spreadsheet via Excel sync → Produces a draft quote with line items + total → Sends to sales rep for review

Pricing changes happen in the spreadsheet, like they always did. The agent's quotes automatically reflect current pricing because it's reading from live sync.

Customer C — a 200-person services firm. Their internal capacity planning is a spreadsheet — who's available, who's overbooked, who's between projects.

They built an agent that monitors inbound staffing requests and matches them against the capacity sheet. The agent suggests assignments. Staffing manager approves with one click.

Spreadsheet stays the system of record. Agent provides the matching intelligence on top.

Why this matters more than we expected

The reason Excel sync turned out to be central:

It removes the migration barrier. Most "AI for your business" platforms require you to migrate data into their system. That's a non-starter for small/mid businesses whose data lives in spreadsheets they don't want to abandon.

Excel sync means you don't migrate. The spreadsheet IS the database now.

It respects existing workflows. People who built and maintain spreadsheets often have deep domain knowledge about HOW the spreadsheet works. Forcing them to migrate to a different system means losing that knowledge. Excel sync preserves it.

It's a wedge into the company. A team that adopts Avery NXR via Excel sync starts with one agent + one spreadsheet. As they get comfortable, they add more agents touching the same spreadsheet. Then more spreadsheets. Then other data sources. Adoption compounds naturally.

It demystifies AI for spreadsheet-driven cultures. A lot of business operations runs on spreadsheets. People who run those operations are skeptical of new AI tools that don't understand their data. Excel sync says: bring your existing spreadsheets, AI works with them.

What we did wrong initially

We treated Excel sync as a technical capability instead of as a strategic feature.

→ Our website mentioned it but didn't lead with it → Our templates didn't show off how to use it → Our docs treated it as an advanced topic instead of a starting point

Customer feedback corrected us. The teams getting the most value out of Avery NXR were the ones leveraging Excel sync heavily.

We've reordered. Excel sync is now featured prominently. We're building templates that show how to use it. Documentation has a dedicated guide.

The lesson for builders

If you're building product, watch what your customers actually USE — not what your marketing emphasizes.

The features customers gravitate toward tell you what your product is actually solving for them. Sometimes it's different from what you thought.

In our case: we thought we were building "an AI agent platform that happens to support Excel."

Customers showed us we'd built "an AI agent platform where Excel data becomes operational."

The reframe matters because it changes positioning, marketing, roadmap.

What's next for Excel sync

Based on customer usage, we're working on:

→ Better conflict resolution UI when human edits and agent edits collide → Validation rule support across more Excel features → Multi-file sync (treating multiple sheets as a coherent dataset) → Google Sheets equivalent (this is a frequent request)

The feature that started as a side capability is now actively shaping our roadmap.

The bigger principle

For most small and mid businesses in 2026, the system of record isn't a database. It's a spreadsheet.

Any AI platform that can't work with spreadsheets natively is forcing migration as a precondition. That precondition is a moat against adoption.

Avery NXR removes the moat. Spreadsheets become operational data. AI agents work against them. Migration not required.

If you have data in spreadsheets that you wish your AI tools could use without migration — try Avery NXR's Excel sync. The capability might solve a problem you didn't realize you'd given up on.

→ avery.software — Free Desktop tier. Drop in your spreadsheet, build an agent that uses it.