Audit, Blueprint, Build, Run. Fixed scope, published prices, everything in your name.
How it runs
What we do
Start from the team that does the work, or the shape of your business.
Two-way syncs with reconciliation, retries and alerts, in accounts you own.
Systems in production, with the numbers, and the screens behind them.
Customer stories
See the work
Field notes from building automation and AI systems, with sources.
From the blog
Before you start
Data consolidation and reporting
We pull data out of your CRMs, project tools, finance and commerce systems, match it on shared IDs, and give you one dashboard and a weekly summary that everyone trusts. Built in your accounts, at a fixed price.
In short
Data consolidation means pulling data from several tools into one place, matched on a shared ID such as a customer, campaign or project, so it is counted once and reported the same way everywhere. SBD builds it on Airtable, Postgres or a warehouse depending on volume, with a live dashboard, a weekly AI-written summary and alerts, in accounts you own.
Sound familiar
What you get
Every figure links to its source record, every source shows when it last synced, and a short summary says what moved this week.
Margin by client
Sources
Weekly summary AI draft, reviewed
Revenue is up 9% on last quarter, led by renewals. Margin on Harbour & Co fell to 14% as logged hours ran 30% over estimate. Support backlog rose after Monday’s release; 12 tickets are past SLA.
Use cases
The tools differ; the job is the same. Connect, match on a shared ID, keep one clean copy, and show it.
Sales
One pipeline from several CRMsTeams on different CRMs after a merger or by region, with no single forecast.
You getDeduplicated accounts and contacts, one pipeline and forecast, stage definitions mapped across systems.
Delivery
Portfolio view across project toolsDelivery tracked in monday.com, ClickUp, Asana and Jira, depending on the team.
You getEvery project on one board with status, owner, due dates and overdue work, refreshed through the day.
Finance
Multi-entity finance roll-upSeveral companies or regions on separate ledgers, consolidated in a spreadsheet each month.
You getCash, receivables and P&L by entity and in total, mapped to one chart of accounts.
Finance
Margin by client or projectHours, costs and invoices live in different tools, so profitability is a guess.
You getRevenue, time and cost joined on a client or project ID, with margin by month.
Revenue
Sales across every channelOrders from a web store, marketplaces and payments reported separately.
You getDaily revenue, units and refunds by channel and product, reconciled to payouts.
Support
Support performance across desksDifferent teams or brands on different helpdesks with different SLA rules.
You getVolume, response and resolution times and backlog in one view, by team and brand.
Marketing
Campaign results to revenueAd platforms, analytics and the CRM each claim the credit.
You getSpend, leads, pipeline and revenue per campaign, joined on a campaign ID.
Nonprofit
Fundraising, impact and programme workDonations, media coverage and programme delivery tracked in separate tools with no shared view.
You getFunds raised, coverage and project progress per campaign, with a weekly summary for trustees and funders.
Operations
Stock across warehouses and channelsInventory in an ERP, a 3PL portal and the store, never quite agreeing.
You getOne stock position by location, with low-stock and mismatch alerts.
Leadership
The weekly leadership packSomeone spends Monday morning copying numbers from six tools into slides.
You getA dashboard plus a short AI-written summary of what moved and why, sent to Slack or email.
Agencies
Client reporting from every toolEach client report assembled by hand from ads, analytics, CRM and project tools.
You getA report per client that refreshes itself, with commentary drafted for review.
People
Headcount, leave and capacityHR records, leave, timesheets and project allocations held in different places.
You getHeadcount, absence and allocated capacity by team, with a forecast of who is free.
Deduplication workflows
The same company appears in the CRM, the project tool and the ledger under different names. Matching them properly is what makes every total on the board trustworthy.
Four records, three systems
customer_id 10482
One golden record that keeps all four source IDs, so every system can still be traced and updated.
Clean every field before comparing: trim and case names, lower-case emails, format phones to E.164, strip “Ltd” and “Inc”, and reduce websites to a domain.
Join on strong identifiers first: email, company domain, tax or registration number, order or invoice ID.
Score the rest on name similarity, address and phone, with blocking so millions of rows compare in minutes, not days.
High scores merge automatically. The grey zone goes to a review queue where a person confirms or rejects each pair in one click.
Survivorship rules decide which source wins per field: billing details from finance, owner from the CRM, the latest phone number from anywhere.
Every merge is logged and reversible, each golden record keeps the IDs from every source, and new records are checked as they arrive.
Built to run
A dashboard is only as good as the pipes behind it. Every build includes the engineering that keeps numbers right after launch.
Only changed records move after the first backfill, so syncs stay fast and inside API limits as volume grows.
Calls back off and retry on errors and limits. Writes are idempotent, so a rerun never creates duplicates.
Row counts, missing fields and totals are reconciled against each source on every run, and breaks raise an alert.
Every connection shows when it last succeeded. A failed or stale sync alerts the owner before anyone reads a wrong number.
Snapshots keep the past, so you can see what a number was last month and why it changed.
Teams, clients or entities see only their rows. Least-privilege keys, stored in a vault, never in code.
From a few thousand records in Airtable to tens of millions in Postgres or a warehouse, on the same model.
Data model, field mappings and runbooks in your accounts, so your team or ours can run it.
How we build it
The same five steps whatever the tools. Matching is where most of the value is, and where most spreadsheet reports go wrong.
Pull from each source by API or webhook. Where a plan has no API, use scheduled exports or files dropped in a folder.
Give records a shared key: a customer, campaign, project or order ID. Map fields and stages, and merge duplicates.
Keep one clean copy with history. Airtable for smaller volumes, Postgres or a warehouse as it grows.
Build the views people need: a board, filters by team or client, drill-down to the source record.
Send a weekly summary of what changed, and alerts when a number crosses a line. Every figure links back to its source.
Choosing the stack
We recommend one after the audit, based on volume, who edits the data, and who needs to see it. You can start small and move up.
Sources
Not listed? If it has an API, an export or a report it can email, it can be connected.
CRM
HubSpot Salesforce Pipedrive Zoho CRM
AttioProjects
monday.com ClickUp Asana Jira
Notion
LinearFinance
Xero Zoho Books NetSuite
Sage StripeCommerce
WooCommerce Amazon
BigCommerce ShipStationSupport
Zendesk Intercom GorgiasMarketing
Klaviyo Mailchimp
Brevo
TypeformData and BI
Airtable Supabase PostgresPrice
FAQ
Bringing data from several tools into one place, matched on shared IDs, so the same customer, project or order is counted once and reported the same way everywhere. The result is usually a dashboard, a set of reports and a regular summary.
Yes. Accounts and contacts are matched and deduplicated, stages are mapped to one set of definitions, and the pipeline is reported once. The CRMs can keep running as they are, or be merged later.
Fields are standardised first, then records are matched on strong identifiers such as email, domain or tax number, then scored on name, address and phone similarity. Confident matches merge automatically, uncertain ones go to a review queue, and survivorship rules decide which source wins per field. Every merge is logged and can be undone.
Some platforms put API access on higher plans or sell it separately. Scheduled exports, emailed reports or files dropped in a shared folder work instead; the data is slightly less fresh, but the dashboard is the same. Check your plan before anyone quotes.
As often as each source allows and the use needs: every few minutes for operations boards, hourly or daily for management reporting. Webhooks give near real-time updates where a tool supports them.
Airtable suits a handful of sources and tens of thousands of records, and lets the team edit as well as view. Postgres with a custom app suits larger volumes, complex logic or customer-facing views. A warehouse with a BI tool suits analysts who want to query freely.
A dashboard from two or three sources typically takes two to four weeks. Multi-CRM or multi-entity consolidations with deduplication take longer, and the audit gives a fixed figure before work starts.
In accounts you own: your Airtable, your database, your cloud project. We work with least-privilege access and hand over everything at the end.
Builds start at $2,500 for a single-source dashboard, and most multi-source consolidations land between $5,000 and $15,000, at a fixed price agreed after a $500 audit that is credited to the build. Care plans from $50 a month keep the connections running.
Platform limits change. Checked October 2026:
Related: Build vs buy tool · ROI calculator · How to run an automation audit
Start here
On a discovery call we sketch the sources, the shared ID and the board, and tell you which stack fits. No charge and no pitch.
Book a discovery call
Ava AI concierge
Ready
AI voice, powered by ElevenLabs. Ava can open pages and fill the contact form for you; you always press send. How we handle data
