Skip to content
Plotwell

AI-native analytics, no data team required

A spreadsheet in, a dashboard and the takeaways out.

Upload a CSV and Plotwell reads the columns, picks the right charts, builds a clean dashboard, and writes plain-language takeaways. Raw data to understanding in seconds, without a data analyst or a BI tool.

Free to start. Every number is read from your data, never invented.

store-2024.csv

Month

date

Channel

category

Revenue

number

Refund

number
2024-01
Organic
24,700
1.8%
Feb 2024
paid search
$19,780
2.0%
2024-03
Email
11,120
2.1
2024/04
Referral
6,300
2.3%

mixed date formats, currency symbols, a stray percent

Total revenue

0

+18%

Orders

0

Revenue by month

Revenue by channel

Takeaway

Revenue is up 18 percent, but refunds are creeping.

read from your data

The problem

Data sits in spreadsheets, unread.

The numbers you need are already in a file. What went up, what is off, what to look at next. But the answer is buried under a hundred rows, and pulling it out means a pivot table, a chart wizard and an afternoon you do not have.

A proper BI tool would help, if you had a data analyst to model the data, name the metrics and build the dashboard. Most small teams do not. So the spreadsheet gets a glance and a scroll, and the story it is telling goes unread.

73%of small teams keep their key numbers in spreadsheets rather than a BI tool
5+ hoursa week the average operator spends building and updating reports by hand
$30k+a year for a BI seat plus the analyst time to set it up and keep it running

Illustrative figures used to describe the problem Plotwell is built for, not results from a survey we ran. They reflect what small teams commonly report about living in spreadsheets and the cost of the tools meant to replace them.

How it works

Three steps, about twenty seconds.

  1. 01Upload a CSV or spreadsheet

    Drop in the file you already have. A sales export, a list of orders, survey responses, ad spend by campaign. Plotwell reads the header row and the data underneath.

  2. 02Plotwell understands it and builds a dashboard

    It works out which columns are dates, categories and numbers, picks the charts that actually fit, and lays out a clean dashboard. No modelling, no metric setup, no chart wizard.

  3. 03Read the plain-language takeaways

    Next to the charts, Plotwell writes what changed, what looks off and what to look at. Every figure is read from your data, so you can act on it, not guess.

See a dashboard

A real dashboard, built from a sample spreadsheet.

This is not a screenshot. Every chart below is aggregated from a sample CSV in the browser, the same way Plotwell handles your file. Toggle the takeaways, the outliers and the columns it read.

Dashboard

Online store, 2024

A year of revenue for a small online store, split across four channels. Revenue is up over the year, orders track it, and the refund rate is worth watching.

48 rows · 6 columnsstore-2024.csvbuilt with gpt-4.1-mini

Total revenue

$0

+18% first to last

Orders

0

Revenue by month

Summed across every channel.

Revenue by channel

Which channel brings the money in.

Refund rate by month

Mean refund rate. Worth watching.

Revenue share by channel

  • Organic40%
  • Paid Se32%
  • Email18%
  • Referral10%

Outliers

Ad Spend

14,125

At 2024-09, 262% above the usual range.

What it does

Everything a first dashboard needs, and nothing to set up.

Understands any spreadsheet

Plotwell reads your header row and works out which columns are dates, categories, numbers or yes/no, coping with mixed date formats, currency symbols and stray percent signs along the way. No mapping, no setup.

Picks the right charts

A trend gets a line, a breakdown gets bars, a share gets a donut. Plotwell chooses what fits your data instead of making you choose.

Plain-language takeaways

What went up, what looks off, what to look at next, written next to the charts in a sentence you can act on.

Outlier and trend flags

The one month, region or row that sits outside the usual range, computed and flagged.

Export and share

Download the data or send a read-only link to a dashboard, no login required to view.

Model-agnostic

Plotwell routes across leading models through one layer, so a lab change is a config change.

Built AI-native

A model is the engine, not a feature.

No fixed rule can look at a spreadsheet it has never seen and know that Region is a category, MRR is money and Signed is a date, then decide the story is revenue by month with churn creeping under it. That is language work, and it is why Plotwell is built on language models rather than a chart template with AI sprinkled on top.

Plotwell is model-agnostic on purpose. Choosing charts and writing takeaways run on a balanced model, quick triage on a cheap one, and wide files escalate to a frontier model. A lab changing price or policy is a configuration change here, not a rebuild.

Understands the columns

Infers each column's type and role from its values, not just its header, so it works on a file made by hand.

Chooses the charts

Reasons about which views answer a real question, then Plotwell aggregates your rows to draw them.

Writes grounded takeaways

Describes direction and change in plain language, and drops any figure it cannot read straight off your data.

The honest part

It describes your data. It never invents numbers.

This is the part we are strict about. Every chart is built by aggregating your actual rows, in code, not by the model. The model chooses which columns to chart and writes the takeaways, and any takeaway that cites a figure Plotwell cannot read straight off your data is dropped before you ever see it.

So a takeaway might say revenue is rising or refunds are creeping, and where it puts a number on that, the number is real. Read the Care page for exactly how.

What Plotwell does

  • Computes every chart by aggregating your real rows
  • Flags outliers it finds with a statistical test, not a guess
  • Writes takeaways about direction and change in plain language
  • Marks every insight as generated

What it never does

  • Invent a total, average or percentage not in your data
  • Let the model type numbers straight onto a chart
  • Train on your data or sell it

In use

For the person everyone forwards the spreadsheet to.

I run ops for a 14-person company and I am the closest thing we have to a data team. Plotwell turns our weekly sales export into the dashboard I used to spend an hour rebuilding, and the takeaways catch things I would have scrolled past.
DPDanielle PryorHead of Operations, Marlow Goods
We were paying for a BI tool nobody could set up. I uploaded three months of ad spend to Plotwell and had a dashboard with the outliers flagged before the BI onboarding call even started. It reads the columns the way a person would.
SOSam OkaforFounder, Northline Coffee

Illustrative customers, used to show who Plotwell is for.

Pricing

A few dashboards free, every month

Prices are in US dollars and that is the only price list there is. No setup fee, no minimum term, cancel from Billing.

MonthlyAnnual

Switch to annual and two months are free.

Free

A few dashboards a month, built properly. Enough to see what Plotwell does with your own numbers.

$0forever

No card required

A few dashboards a month, files up to 5,000 rows

  • 4 dashboards a month
  • Files up to 5,000 rows
  • The charts that fit your data, chosen for you
  • Plain-language takeaways and outlier flags
  • Upload a CSV or a spreadsheet
Most popular

Pro

For the operator who lives in spreadsheets. Unlimited dashboards, larger files, export and a share link.

$19per month, USD

$228 a year, or $192 on annual

Start on Pro

Card payments are not switched on in this deployment yet, so this joins the list and we email you the moment it is.

Unlimited dashboards, files up to 50,000 rows, export and share

  • Everything in Free
  • Unlimited dashboards
  • Files up to 50,000 rows
  • Export a dashboard as an image or the data as CSV
  • Share a read-only link to any dashboard
  • Dashboard history you can come back to
  • Email support

Team

For the team that forwards spreadsheets to one person. Five seats, one shared workspace, answers the same day.

$49per month, USD

$588 a year, or $492 on annual

Start on Team

Card payments are not switched on in this deployment yet, so this joins the list and we email you the moment it is.

5 seats, shared workspace, priority support

  • Everything in Pro
  • 5 seats
  • A shared workspace the whole team can open
  • Dashboards the team can find in one place
  • Priority support with a same-day first reply
  • Invoiced billing on request

Questions

The ones people ask first

If yours is not here, send it to us and a person will answer.

Your next spreadsheet, understood.

Upload a CSV and get a clean dashboard with plain-language takeaways in seconds. A few free every month, no card.