Store first, then push
Every enquiry saved to your own database before being sent anywhere. If the CRM is down or the API changes, the lead still exists. This single decision prevents the most expensive failure in this category.
The most expensive integration failure is the silent one. Enquiries submit successfully, never arrive anywhere, and nobody notices for weeks.
Connecting a website to a CRM is straightforward. Making it reliable is not. The work that matters is deciding what happens when the other system is unavailable, storing every submission on your own server before pushing it anywhere, and mapping fields between two systems that describe the same thing differently.
Every enquiry saved to your own database before being sent anywhere. If the CRM is down or the API changes, the lead still exists. This single decision prevents the most expensive failure in this category.
What happens when the other system rejects a record or is unavailable: queue and retry, alert someone, or fail silently. Most integrations do the last one by default, which is how enquiries vanish.
The cost is not the connection, it is reconciling two systems that model the same information differently. One has a single name field, the other has three. One requires a phone number, the other does not.
What happens when the same person enquires twice. Without a rule you get duplicate records, and sales teams stop trusting the CRM, which is worse than not having the integration.
A native connector, then middleware such as Zapier or Make, then custom code. Teams commission custom work for things a connector already handles, which is expensive and unnecessary.
APIs change, credentials expire, endpoints get deprecated. Someone should be alerted when records stop flowing, rather than finding out from a sales team wondering where the leads went.
| Problem | How you find out |
|---|---|
| No local storage before pushing | Weeks later, when leads are missing |
| No failure alerting | From a salesperson, eventually |
| No duplicate rule | When the CRM stops being trusted |
| Credentials expired | Silently, until someone checks |
| API version deprecated | When it stops working entirely |
| Nobody owns it | At the worst possible moment |
| Ask them | Good answer | Walk away if |
|---|---|---|
| Who owns the accounts and data? | You do, in your own name | Accounts in the agency's name |
| Can I open three live examples? | URLs you can check yourself | Screenshots and a PDF |
| What is explicitly not included? | A written list with change pricing | Everything is included |
| What happens when it breaks? | A named process and response time | Not discussed |
The failure question matters most on anything connected to another system, because integrations fail quietly.
Silent failure. The form submits successfully, the record never reaches the CRM, and nobody notices for weeks. Storing every submission in your own database before pushing it anywhere prevents the loss entirely.
For low volume and simple flows, often yes, and it is far cheaper than custom work. It becomes a poor fit at high volume where per-task pricing adds up, or where you need reliable retries and error handling.
Because the cost is not the connection, it is reconciling two systems that describe the same information differently, and deciding what happens when data fails validation. Ask for that mapping work to be scoped explicitly rather than assumed.
That depends entirely on how the integration was built. A well-built one queues and retries; a naive one loses the enquiry. This should be a decision made at the start, not something discovered later.
The common ones including Zoho, HubSpot, Salesforce, Freshsales and LeadSquared, plus anything with a documented API. Tell us what you run before we quote so integration is scoped rather than assumed.
Someone must. APIs change, credentials expire and endpoints get deprecated. Agree ownership at the start, because unmaintained integrations fail quietly and the failure is discovered through missing data.
Send us what you are trying to fix and we will tell you what it takes, what it costs, and whether we are the right people for it. If we are not, we will say so.