Hi everyone. I've been using Anyword for a few months now to generate marketing copy for different campaigns, and it's been super helpful. But I've run into a bit of a workflow snag.
I often generate dozens of variations for a single project, tagging the ones I like. When it's time to move everything over to our project management tool or a spreadsheet for the team to review, I find myself having to open each individual piece of copy and copy-paste it out. It's really time-consuming.
Is there a way to export *all* the generated copy for a project, or at least a batch of selected variations, at once? Like a CSV or JSON export? I've poked around the dashboard and couldn't find an obvious "Export All" button.
If there isn't a built-in way, has anyone figured out a workaround? I was thinking maybe there's an API endpoint I could call with a simple Python script? Something like this would be ideal, if it exists:
```python
# Pseudo-code wishlist
import requests
response = requests.get('https://api.anyword.com/v1/projects/{project_id}/copy',
headers={'Authorization': 'Bearer API_KEY'})
# Get a JSON array of all my variants
```
Any tips would be really appreciated. This seems like a common need for anyone doing analysis on what copy performs best.
The API approach you're considering is definitely the right direction for automation, but I need to temper your expectations a bit. After reviewing their public documentation, Anyword's current API appears to be focused on generation and prediction calls, not bulk retrieval of your historical generated content. The `/generate` and `/predict` endpoints are well documented, but a `/projects/{id}/copy` endpoint, as you've hypothesized, isn't listed.
Your workaround likely needs to be a two-step process. First, you'd need to use their API to fetch your project or campaign list to get the correct identifiers. Then, you'd have to iterate through each item and potentially call a detailed endpoint per piece of copy, assuming it exists. This could still hit rate limits and be complex.
A more immediate, though still manual, solution might be to use your browser's developer tools while on your project dashboard. Monitor the network tab as you scroll or load your variants; the front-end is likely fetching your copy via *some* internal API call. You could potentially find that endpoint and replicate it in a script. The data format will probably be JSON, which you could then parse into CSV.
If you go the manual copy-paste route as a last resort, at least use a multi-cursor editor or a browser extension that lets you select multiple non-contiguous text blocks at once. It's still manual, but slightly less painful than one-by-one clicking.
I feel your pain on the copy-paste marathon! I hit the same wall last quarter and the built-in export button is... well, it's not there.
Your API hunch is a good instinct, but I checked with our CS rep recently. They confirmed there isn't an endpoint for bulk copy retrieval yet. The API is really just for generating new content. So that Python script sadly won't work right now.
My temporary fix was using a browser extension for data scraping. It's clunky, but I was able to pull all the text on a project page into a sheet in a few clicks. Not elegant, but faster than doing it all by hand!
Happy customers, happy life.
You've hit on a really common pain point with these platforms. Your pseudo-code perfectly captures the ideal solution, and it's a feature request I've passed along to a couple of vendors myself.
Since the API route isn't ready yet, and browser extensions can be hit-or-miss with dynamic content, you might consider a slightly different manual method that's saved me time. When I'm on the project page with all my variants, I'll select all the text on the page (Ctrl+A or Cmd+A) and paste it into a basic text editor. This often grabs all the copy in one go, though you'll get some extra UI text you need to clean out. A quick find-and-replace for the site's page labels can get you a clean list faster than clicking into each item.
Has anyone else found a reliable scraper that works with the latest Anyword interface?
That's a clever idea to check the network tab. You're right, the frontend has to get the data from somewhere.
I've tried that method before on a different platform. The tricky part is often the authentication - those internal calls usually require a session token or specific headers that can be a pain to replicate in a standalone script. Plus, they can change without warning since it's not a public API.
Have you actually been able to reverse-engineer an internal endpoint from a tool like this into a working, repeatable script? In my experience, the juice isn't always worth the squeeze compared to a cleaner manual workaround.