Google Maps Scraper

Hey, thanks for the report, we will have a look asap.

Hello,

Thanks a lot for the report, we will resolve it today.

1. The problem for this run GQQKdEh2DVe2ytMBq is that you have hamburg as the first search term for which Google just shows the city of Hamburg, nothing else. This feels like a typo on your end? This was previously handled well but with new handling slipped back to repeated scrape. I notified the developer to handle this gracefully, e.g. just output the single place and throw away the scanning.

Hello Lukáš,

Thank you for reverting back to me quickly.

1. Indeed you're right: on the following 6 runs I made a typo of adding Hamburg: GQQKdEh2DVe2ytMBq / oQkJC7xj4NHygurnp / RkVVFtzuOHbdPGIDh / bSxNyz3TPPmZNtyAP / SrpsBp0qXUMpVtnuD / lz6KZwcJ5lUTvue8c.

For these 6, do you confirm simply re-launching them from scratch would now work, should I remove the "hamburg" mention from search terms? (no need for a compensation given that it was my mistake) not sure the scraper works though, given my following comment.

For the other runs, I also see the mention of "Finishing scroll - didn't find the cursor for the next page" in their log, (e.g. lntm3rFt0BlUg1M4B) despite no typo in the inputted search terms. On other past runs, I didn't come across such mention in their log (e.g. jMtwHxzP1yJCsor4W).

Could you confirm what went wrong?

Thank you.

Hey,

1. Please wait with the launch until the dev investigates the other point. But yes please remove the hamburg search term and keep the other real terms.

We will get back to you asap today for the other problems.

Coming up with the resolution of the issue.

The developer analyzed the runs and there is no bug, just the log message is written in a confusing way which will be fixed. Basically, the scraper splits the area into "map chunks" and for each of the chunks it uses the "Search this area" feature to only get places at that location to increase the capture rate. Unless you are searching in cities or for very common terms, most areas don't contain anything which is why it seems like the scraper is stuck, it still finds something once in a while though. For some runs you shared, the $/place ratio is quite good but for some, it is not ideal - this depends mostly on the density of the searched area and how frequent the type of the place is.

We are working on a more optimized approach that will start on lower zoom and zoom in only when needed but it is not yet ready.

Would you expect very different number of places for some of the searches you did? We can explore these manually.

Hello back Lukáš , thank you very much for your help here, very clear and nothing to add on my end. I will re-run the typo-ed ones & keep the other ones as they are then. Thank you and please do ping me whenever that update is done, I'd love to give it a try on past scraped runs.