Imagine if a technology existed that could read your email for you, and use that data to trigger a workflow, or automate something. The good news is that it already exists, and has for some time. It’s called an email parser. The bad news is that it is at best kind-of helpful, and at worst, wildly inaccurate.
Although email is rapidly becoming an outdated tool, it is still widely used and email parsing does have a place in the automation toolbox.
WHAT ARE EMAIL PARSERS?
The way email parsers work is that they use the structure of an email (usually text or HTML) to look for matches to pre-defined fields. The two email parsers I would usually recommend are Zapier Email Parser, and Mailparser.
So, using the Zapier parser as an example, you set up a template by highlighting and naming the data that you want to look for in the email. Then, by some dark magic (actually, some code) the email parser will return the data it has found in usable format (e.g. for a spreadsheet, CRM, etc). Like this:
WHEN IT WORKS AND WHEN IT DOESN’T
It works very well when you are receiving structurally consistent emails, like order confirmations, or even contact form submissions from your website. A lot of use cases revolve around notification emails from older websites. So, in summary it works from machine to machine very effectively.
Except when it doesn’t.
Even if the email is coming from a machine, format can still vary considerably, and that can confuse the robots that run the parser. So what you get at the other end may not be what you’re expecting.
It gets even harder when there are attachments to the email (possible, using Mailparser), but still heavily reliant on the content of the attachments.
ENTER, HUMANS.
Humans are the most unpredictable factor to bring into the mix. We say things in different ways, we all have differing language abilities, dialects, and writing style. English, in particular, is quite idiosyncratic. Take this sentence for example:
I saw a man on a hill with a telescope
What does that mean to you? It could mean:
- There’s a man on a hill, and I’m watching him with my telescope.
- There’s a man on a hill, who I’m seeing, and he has a telescope.
- There’s a man, and he’s on a hill that also has a telescope on it.
- I’m on a hill, and I saw a man using a telescope.
- There’s a man on a hill, and I’m sawing him with a telescope.
If an email parser is looking for particular phrases, or intent, it may come back with data that is rather peculiar.
EMAIL PARSER OPTIONS
Usually, I would say that an email parser is a tool of last resort. It is used when there are absolutely no other options available. Generally this is when you are not in control of the source of information. A typical example would be extracting information from websites that have no other method of doing so, no API, and are hard to use a crawler on (like a real estate listings site, for example).
The most viable options for an email parser include:
- API or other programmatic integration method between the data source and your system/s
- Zapier
- Website crawling (Apify is my pick)
- Website forms
- Chatbots
Naturally, the options that work for you will depend on the use case.
THE FINAL VERDICT ON EMAIL PARSERS
Avoid email parsers when you can. They are a kind of duct-tape automation tool that can be risky and unpredictable.
When you have to use them, there are some great options on the market that do it really well. Mailparser offers a feature-rich experience with a very well thought through product. Zapier’s mail parsing tool is a great basic option with an easy to use interface.
You can learn more about business automation with our Education Section here.