HubSpot + MIE Trak Pro integration

MIE Trak Pro began as an estimator’s tool. Its founder wrote a quoting program in 1984 for his father’s sheet metal shop, and the ERP that grew from it is still organized around the RFQ, the quote and the work order. It serves sheet metal fabricators, precision engineers and other discrete manufacturers. For the people selling a job shop’s capacity, that design means calling the office to learn which RFQs are still open and which customers have gone quiet.

MIE has met part of that need itself. Its knowledge centre documents a HubSpot service and a Salesforce service, beside MIE CRM, a web application of its own that reads the MIE Trak Pro database. The HubSpot service is deliberately narrow: customer details, one way, into HubSpot contacts. That builds a marketing list, not a pipeline.

The gap is the part of a job shop that decides revenue. RFQs waiting on approval, sales orders in production and invoices posted but not yet paid all live in MIE Trak Pro’s SQL Server database. A sales team that works in HubSpot needs them on the company record, on a schedule it can trust, without a seat in the ERP or a spreadsheet from accounting.

Five objects carry a job shop’s pursuit. Customer parties become companies, the contacts on each party become contacts, RFQs become deals, and sales orders and posted invoices land on custom objects associated to the company.

MIE Trak Pro calls every relationship a party, and one party can hold several types at once: a firm can be a customer and a supplier in the same record. The customer flag decides what reaches HubSpot, or suppliers turn up in the sales pipeline.

The RFQ is the deal because it is the record a customer accepts or declines, and its statuses, from draft through ready for approval, approved and submitted to closed, map onto deal stages without invention. Invoices sync only once posted. An unconfirmed invoice is still a draft, and even a confirmed one can be unconfirmed to change a shipped quantity, so neither belongs in a sales report.

Every key is the MIE Trak Pro identifier, held in a custom HubSpot property. Company names change as the office tidies records. Identifiers do not.

MIE Trak Pro reaches HubSpot through a managed sync that reads SQL Server. It requires Microsoft SQL Server 2017 to 2022, Standard or Enterprise, on Windows Server, and MIE’s own integrations read that database directly: its HubSpot and Salesforce services take a SQL connection string, and its shipping integrations connect through an ODBC data source with a SQL login.

We define read-only views over the tables each object needs and give the sync its own read-only SQL login. A Skyvia agent inside your network reads the views without any inbound port, and a chain of packages, run nightly, upserts companies first, then contacts, deals, orders and invoices, each record looked up on its MIE Trak Pro key before it is written. MIE publishes no modified date for these tables, and its Salesforce service tracks edited parties with an internal sync flag rather than a timestamp, so each view reloads in full and the upsert updates records in place. Where MIE hosts the system on AWS, the agent runs beside that SQL Server, and MIE’s hosting guide notes that firewall and port settings may need your IT team. A failed package is rerun from where it stopped.

Start with what MIE already ships. The HubSpot service installs on your server, reads the database every fifteen minutes by default and copies each customer party that has an email address into a HubSpot contact. If a customer list for marketing is all you need, it is the first thing to try.

It stops there. It sends no RFQs, orders or invoices, and it puts a company’s details on a contact. The managed sync covers those, one way, from read-only views. A custom integration earns its place only when HubSpot has to write back, a won deal becoming a sales order, and then it should produce the CSV file MIE’s electronic sales order import already accepts.

1. Scope. We find out who hosts the SQL Server, whether MIE’s HubSpot service is already running, and which RFQ statuses sales treats as open.
2. Architect. We write the view contract, the keys, the load order and a decision on the vendor service: retired, or kept for contacts alone.
3. Build. We point the views at a training database, which MIE’s documentation says many customers keep, and the packages at a HubSpot sandbox.
4. Validate. We check a full sandbox load against an MIE Trak Pro report of open RFQs and posted invoices, and the office manager signs off every difference.
5. Operate. We record in the runbook whether MIE’s own service still runs, read the chain’s history each morning, and retest the views after every MIE Trak Pro update.