This past Tuesday, our awesome Developer Communications Director, Adam DuVander, and Developer Evangelist, Kunal Batra, hosted our monthly webcast Be an Email Switch Hitter: Inside SendGrid’s Parse Webhook. We ran the bases starting with the basics on the differences between an API and a webhook, to actual use cases from customers and hackathon demos, to a live demo on how to actually integrate the Parse Webhook into your email program.
A quick review for those of you who aren’t familiar, SendGrid’s Parse Webhook allows your web application to receive email. Why would that matter? Just imagine the last time you received an email from an organization and read that glaring “no-reply” from address. It’s a little off-putting. Nobody wants to feel like they are being spoken to instead of spoken with. With the Parse Webhook you can open up the lines of communication and give recipients the chance to respond!
As promised, here’s the follow-up post that covers the live Q&A portion of the webcast.
Why use the Parse Webhook when you could just use IMAP on an inbox?
You could set up an inbox and just have responses go there, but then you would have to poll that inbox regularly to find out if there is anything new. The Parse Webhook sends it to you as soon as it’s available, allowing you to automate the process.
What is the max file size attachments the Parse Webhook can handle?
Altogether, it depends on your own ISP, but the Parse Webhook can handle 20 megabytes total (the email, its contents, plus any attachments).
How can I use firstname.lastname@example.org and not have to route all my inbound email through SendGrid?
So to be able to use the Parse Webhook, you have to be able to route through all of the email at a domain or subdomain. One solution is to use a subdomain and have people email support@(something).(your domain). If you already have the support@ email address, then you can forward anything to that email to an address on the sub-domain you use for the Parse Webhook.
What’s the difference between the Parse Webhook and the Event Webhook?
The Event Webhook sends nine different events (delivered, opened, bounced, etc.) as they happen. So as soon as an email is delivered, or if it bounces or someone unsubscribes, these pieces of data (that SendGrid is collecting on behalf of your sending) then can immediately forward on to a URL you provided.
Does SendGrid do OCR? What app was used for the fun example?
We do not do OCR. It’s just another API we use to communicate with the SendGrid API. You can use any sort of free OCR library and hook it up with SendGrid, and that way you can just send attachments in and use the API or library to extract text.
What would happen if an HTTP post is made by the Parse Webhook and the HTTP endpoint is temporarily unavailable? Does it retry?
Yes it does retry. Unless it gets the 200 OK message, it keeps retrying for up to three days at certain intervals. Then, if it can’t be delivered after three days it would be dropped.
Do your Webhooks support HTTPS?
What happens to incoming email marked as spam when you check the feature to “check incoming emails for spam?”
When you select the “check incoming emails for spam” checkbox, it’s actually just generating spam reports and spam scores for those emails. You will receive all the emails, then your application decides what to do based upon that spam score. So you can decide “I don’t want to see anything below a spam score five, or three” or however tight you want to make your application.
How do you differentiate between the new content of the email vs. the quoted reply text?
We just send you the entire body text to your application and from there you can go ahead and add that logic between “this is new content” and “this is a reply” based on the count size or however else the formatting is. So, the Parse Webhook is sending you the body as is.