Key takeaways
- MGID is a native ad network: your ad appears as a recommended link in a widget on a publisher's site, in the same style as the content around it.
- It sits in the same family as Taboola and Outbrain, so the same logic applies: sell the click with an editorial-style story, then convert on a page that continues it.
- Meta-style product ads usually underperform here. Story-led headlines, images and landing pages are what the format rewards.
- Start with a narrow test: one offer, one market, one device, a handful of creatives, and a clear definition of success set beforehand.
- Judge MGID by your own cost per acquisition and payback, not by click-through rate alone.
If you run paid social, you already know how to buy attention from people who were looking at something else. So what is MGID, and why does it keep coming up when brands look for places to grow beyond Meta? MGID is a native advertising network. It puts your ads inside recommendation widgets on publisher websites, where they look like the articles around them. This guide explains how it works, which parts matter for a DTC brand, how a first test could look, and the mistakes that waste budget. It assumes you know Meta ads but not native.
What MGID is, and what it is not
MGID is a native advertising network. It places your ads as recommended content links inside widgets on publisher websites, matching the look of the page around them. Advertisers pay to send readers to a landing page, article or product page. It works like Taboola and Outbrain, with its own publisher inventory and tools.
A native ad takes the form of the content it sits next to. On a news or lifestyle site, you will see a block under the article with headings like "You may also like" or "Around the web." Each tile has a picture and a headline. Some of those tiles are ads. MGID is one of the networks that sells and serves those tiles.
The network connects two sides. Publishers add MGID widgets to their pages to earn money. Advertisers like you upload ads and pay to have them shown in those widgets. MGID runs the auction between advertisers and decides which tile appears where.
It is not a social platform, so there is no profile data like the interest and lookalike audiences you are used to in Meta. It is not search either. People are not typing a need into a box. They are reading something and your tile catches their eye. That difference shapes every decision that follows.
For the wider picture of the format, see the complete guide to native advertising. For a comparison of networks, see our rundown of native ad networks for e-commerce brands.
How MGID works
The basic flow
- You create a campaign in the MGID platform and choose a goal, a destination URL, countries, devices and a budget.
- You upload ads. Each ad is usually an image, a headline and sometimes a brand name.
- MGID runs an auction each time a widget loads. Your bid, plus signals about how well your ad is likely to perform, decide whether you appear.
- A reader clicks your tile and lands on your page. You are charged according to the buying model you chose, which for most native campaigns is per click. Check the options in your account.
- You track what happens next with a pixel or postback (a small piece of code or a server call that tells the platform when a sale happens) and adjust bids, sources and creatives.
Why the click is different from a Meta click
On Meta, the ad interrupts a feed that people scroll for its own sake. In a native widget, the reader has just finished or abandoned an article and is browsing for the next thing to read. Your tile competes with editorial headlines, not with other ads. A tile that looks like a product catalogue entry stands out in the wrong way. One that promises a useful or surprising story tends to get the click.
That is why many DTC brands send native traffic to an advertorial: a landing page written like an article that explains the problem, then introduces the product. We cover that format in our guide to advertorials for DTC.
The parts that matter to a DTC brand
Creative: headline and image do the targeting
With fewer audience tools than Meta, the headline and image decide who clicks. A headline about a specific problem ("Why your running shoes feel flat after a few months") draws the right reader. A generic brand claim draws casual clicks that do not buy. Test several angles per offer, not several colours of one angle.
Landing page: it must continue the story
If the tile promises a story and the click lands on a plain product grid, readers leave. The page should pick up the same promise, build the case, then lead to the product page. Keep the headline and first image consistent with the tile.
Placement and device
Results differ by publisher, placement and device. Most platforms let you see performance by source and block or bid down the weak ones. A page that works on a desktop article can feel cramped on a phone, so read our guide to mobile native advertising before you build.
Measurement: judge by your own numbers
Native traffic often sits earlier in the buying journey than Meta retargeting, so it may convert more slowly. Look at cost per acquisition and payback over a window you set in advance, and compare with Meta on the same basis. Click-through rate alone will mislead you because curiosity clicks are cheap and rarely buy.
| Meta | MGID (native) | |
|---|---|---|
| Where the ad appears | Inside a social feed | In a widget below or beside publisher content |
| How people are found | Interest, behaviour and lookalike audiences | Mostly by context, geography, device and the source site |
| What carries the ad | Image or video with copy | Headline and thumbnail |
| Typical landing page | Product or collection page | Often an advertorial or article-style page first |
| Main lever for improvement | Audience and creative | Creative, source-level bids and landing page |
How dependent is your brand on Meta?Get your Native Ads Readiness Score and Platform Dependency Score in 2 minutes. It is free.
Take the testA worked example: a first MGID test
The example below uses a made-up product, not a client case: a posture-support cushion for office chairs, sold from a Shopify store. The brand already runs Meta and wants to see whether native can add a second channel.
- Set the success rule first. The brand decides the cost per acquisition it can accept and how long it will wait for a read. It writes this down before spending.
- Pick one market and one device. The brand chooses its best-performing country and mobile only. Limiting the test to one market and one device gives a cleaner read on what is working.
- Build one advertorial. A short article-style page: the problem of back pain after long desk hours, why common fixes fall short, how the cushion is designed, a few customer reviews, then a button to the product page.
- Write several tiles for the same offer. Each uses a different angle, such as a back-pain headline, a "desk setup mistake" headline and a comparison headline. Each gets a different real photo, not a stock product shot on white.
- Launch with a controlled budget. Run all ads at the same bid so the platform has data on each. Keep the budget small enough that you can afford to be wrong, but large enough to gather clicks on every ad.
- Install tracking and check it. Place a test order and confirm the sale is reported back to MGID and to your own analytics.
- Review by source and by ad. After the review window, pause ads and sources that spend without a sale. Raise bids on the ones that convert, and write new tiles in the style of the winners.
- Decide against the rule from step 1. If the cost per acquisition is inside your limit, widen to another country or device. If not, change the creative or page before changing the channel.
For budget planning, our post on how much native advertising costs explains what drives spend, and for a comparison with the largest network, read our guide on running Taboola ads.
Common mistakes
- Reusing Meta creative as is. Polished product ads with price badges look out of place in a content widget and get ignored or clicked by the wrong people.
- Sending traffic to the homepage. A cold reader needs a reason to care. Use an article-style page or a focused product page that matches the tile.
- Judging too early or on the wrong metric. A low click cost means little if no one buys. Decide the review window and metric in advance.
- Testing everything at once. Several countries, devices and offers in one campaign make it impossible to see what worked.
- Leaving every source on. Some sites will send traffic that never converts. Review performance by source and cut the weak ones.
- Skipping tracking checks. If sales are not reported properly, the platform cannot optimise and you cannot judge the result.
- Making claims you cannot support. Exaggerated health or results claims can get ads rejected and hurt trust. Keep copy honest and specific.
What to do next
Decide whether native fits your business before you commit budget. It tends to suit products with a story to tell, a problem to explain or a clear before-and-after. It is a harder fit for pure impulse items with nothing to say.
If it does fit, build one advertorial and a small set of tiles, then run the narrow test above. If you plan to compare networks, our Taboola vs Outbrain guide covers how to structure a fair test, and the same logic applies when you add MGID.
If you would rather not build and run this yourself, a native advertising agency can handle the creative, setup and optimisation. Whoever does it, hold them to the same rule: a cost per acquisition and a review window agreed before spend starts.
Frequently asked questions
What is MGID?
MGID is a native advertising network. It shows your ads as recommended content tiles in widgets on publisher websites, styled to match the page. Advertisers use it to send readers to articles, advertorials or product pages.
Is MGID the same as Taboola or Outbrain?
No, but they work on the same basic model. Each is a separate network with its own publisher partners, tools and auction. Brands often run more than one and compare results. See our comparison of Taboola and Outbrain for how to test them side by side.
Does MGID work for e-commerce brands?
It can, especially for products that benefit from explanation, such as health, home, personal care or problem-solving items. Results depend on your creative, landing page and offer, so run a controlled test and judge by your own cost per acquisition.
How does MGID charge advertisers?
Native networks commonly charge per click, though the exact options depend on the platform and can change. Check the buying models available in your MGID account and compare them against your target cost per acquisition.
What do I need to start advertising on MGID?
You need an account, a landing page that matches your ads, working conversion tracking, several headline and image combinations, and a clear success rule. An article-style page usually converts cold native traffic better than a plain store page.
The bottom line
MGID is another way to reach people while they read, using headlines and thumbnails instead of audience targeting. For a DTC brand, the work sits in the story: tiles that attract the right reader and a page that continues what the tile promised. Start with a narrow test, write down what success means before you spend, and let your own numbers decide whether MGID deserves a bigger share of budget.
