MaskWright

2026-06-26 · Morgan Ellis · 835 words

Instagram public research versus a scrape

2023 code guide is the weak competitor page. Public research versus scrape. No code stealth. Look with a cold room, or use official channels.

Instagram public research versus a scrape 1

The 2023 code guides treat Instagram as a scrape target. Endpoints, pagination, a headed Chrome when the endpoint closes. That is the weak competitor page. This commercial URL is the fork: public research versus a scrape. Look with a cold profile, or use official channels. No code stealth.

What web scraping is already named the doors. Instagram's public grid is not a dataset because a gist said so.

Public research

A person opens a public profile or a public Reel in a folder that holds no brand login, no personal Instagram, no Business Suite cookie. They take a note. They screenshot a public creative. They close the profile.

That is research. Instagram content research without mixing jars is the isolation sibling on the social desk. Public pages, slowly is the pace. robots.txt and terms come first. If the client path is disallowed, the person can still look, and a program should not harvest.

Official Meta tools exist when the job is your own account: insights, exports, partner APIs you were approved for. Official APIs versus headed collection. Instagram scheduling without unofficial helpers is the scheduler fork. Suite is not a scrape client.

A scrape

A program that pages public (or "almost public") HTML, parses captions and graphs, and writes a lead file. A login so the graph is larger. A residential exit so the host sees many people. A Chrome disguise so the client looks headed. That is a scraping browser with a harvest job. What a scraping browser is. We will not hide it. We will not hide a scraper.

That is the 2023 guide. I will not refresh the selectors. I will not name unofficial endpoints. The Python overview will not grow an Instagram chapter. Web scraping with Python, without a stealth stack.

Behind-login collection is not research. A cookie paste into a collector is a live key. If the leftover is people you could message, you are in a personal-data job we skip.

Official insights on an account you administer are not a scrape. They are a dashboard you already have a right to open, in the brand profile, with a Page role. Export from that dashboard if Meta offers a button. Do not parse the public grid to rebuild insights you were not given.

JobMethodScrape
Three public hooks this weekCold profile, notesNo
Insights on an account you adminOfficial dashboard / exportNo
Follower graph you were not givenStopYes, and refused
Caption archive for six monthsOfficial door or stopYes, if it is a paging client

The cold profile on Windows

Empty cookie store. No Suite. No ads. Open the public URL. Save notes outside maskwright-data. Close the profile before you open the brand profile.

Cookie hygiene is for authorized sessions you already have a right to hold. Do not import those cookies into a research client.

An exit on a cold profile is a path you brought, not a hide for a harvest. A product that sells Instagram scrapers fails permission on the first screen. How to judge a web scraping tool.

A captcha or a wall on a paging client is a stop, then a search for the official channel.

LinkedIn is the same ethic, different URL

We will not scrape LinkedIn inboxes. Official API and public pages. Instagram gets this page because people search "instagram scraping" and land on code. The answer does not soften into a "light scrape."

Scraping ethics we follow is the list. Look, official door, or stop.

A competitive question that fits on a sticky note usually fits in a cold-profile afternoon. "What three hooks is this public account using this week?" is a look. "Every caption and every commenter for six months" is a harvest. The first job does not need a script. The second job does not become research because you used Python.

If a client asks for a follower graph, ask which official insight they already have on accounts they administer. If they have none, the graph is not a deliverable we will help collect. Offer a cold-profile look at public creatives, or official Meta tools on profiles they own. Those are the two products. A scrape is not a third.

MaskWright 0.1 isolates the cold profile from the Suite profile. It will not page the grid. Bulk start will not become a profile walker. We will not load an unofficial helper as an unpacked extension to finish the file.

The Scraping hub holds the ethics cluster. This page is only Instagram's fork. Public look in a cold profile. Official channels for your own data.

FAQ

Can I use Python if I go slowly?

A slow honest client still needs permission. If robots or terms disallow the path, a polite interval does not create a door. Most Instagram "research" jobs are a person looking, not a script.

Should I open public Reels in the brand Suite profile?

No. Keep the look cold. Suite is for accounts you administer.

Related notes