How To Build A Media List That Actually Holds Up

Most media lists are a spreadsheet of outlet names and a column for email addresses. That works right up until the moment somebody asks a question the spreadsheet cannot answer, and in our experience that question is almost always the same one: how do you know this outlet is real?

Here's the thing. In Indonesia, roughly 43,000 sites publish news online. Of those, 988 cyber media outlets are recorded by Dewan Pers, the national press council, based on a direct pull from its portal on 2 September 2026. That's about 2.3 percent. If your media list has no column for that distinction, you are carrying risk you cannot see.

This is a working guide to building a media list that holds up when procurement, legal, or a nervous CMO starts asking questions. It's written from the data side rather than the outreach side, because the outreach advice is everywhere already and the data advice mostly isn't.

What a media list is actually for

Two different jobs get called the same thing, and mixing them up causes most of the pain.

A journalist database is a list of people. Names, beats, contact details, what they've written recently. You use it for earned coverage, for pitching, for building relationships over months. Cision and Muck Rack sell this, and it goes stale fast because reporters move.

A media list, in the paid-placement sense, is a list of outlets. Domain, reach, authority, editorial vertical, what formats they accept, whether a link is dofollow or marked sponsored. You use it for distribution and for buying placement.

Both are legitimate. They just answer different questions, and a spreadsheet trying to do both usually does neither well. Pick one job per file. If you need both, keep them in separate tabs with a shared outlet key so you can join them when you need to.

The columns that earn their place

Every media list starts small and grows into a monster. Fifty columns, most of them empty, nobody remembers who added the one called "notes 2". So here's the discipline: a column earns its place only if a decision changes based on what's in it.

By that test, these survive.

The verification layer nobody builds

This is the part that separates a list you can defend from a list you can only apologise for.

In markets with a press council or press registry, there is a public record of which outlets have been checked by somebody other than the outlet itself. Indonesia has Dewan Pers. It publishes two states, and the difference matters more than people assume.

Terverifikasi Administratif means the paperwork passed. Legal entity, responsible editor, registered address, all in order on paper.

Terverifikasi Faktual means somebody from the press council physically visited the newsroom, checked whether the registered address is actually an office, looked at the editorial workflow, and confirmed the outlet publishes on a regular schedule.

Of the 988 cyber outlets recorded, 830 hold factual verification and 158 sit at administrative. So most have been inspected. That's genuinely useful to know when a client in a regulated sector asks what "verified media" means in your proposal.

Two operational details will bite you if you skip them.

First, the press council register lists media names and legal entity names, not domains. There is no URL column. So matching your list against the register is a name-matching exercise, and name matching is where wrong answers come from. Match on the registered PT name against the masthead on the site, not on the brand name. We built this mapping for our own directory and had to label each match by confidence level, because roughly five percent could not be matched with certainty from the name alone.

Second, and this one surprises people, the certificates expire. Factual verification runs five years. Dewan Pers started a data refresh program in October 2025, and roughly 300 outlets were removed from the official register between December 2025 and June 2026 for not renewing. Three hundred, in six months.

Which means a verified-media list you exported last year is probably wrong now. Not slightly wrong. Structurally wrong, in a way that shows up when somebody checks.

Domain Rating is not a quality signal, and treating it as one costs money

Let's be direct about this, because it's the single most common mistake we see in media lists built by performance teams.

Domain Rating, from Ahrefs, measures backlink profile strength on a 0 to 100 logarithmic scale. It is a decent proxy for search visibility. It says nothing about editorial standards, nothing about legal standing, nothing about whether anyone actually reads the site.

From the dataset behind our own directory, the mismatch is easy to demonstrate. There are regional newsrooms sitting below DR 30 that hold full factual verification, meaning a state body walked into their office. There are aggregators above DR 50 with no registration at all. Pick on DR alone and you can end up buying placement on a site nobody has ever checked, while skipping the licensed newsroom your client's compliance team would have preferred.

Here's a rough way to think about it. DR answers "how strong is this domain in search". Verification answers "has an independent body examined this publisher". Reach answers "will humans read it". Three questions, three columns, and no single number covers all three.

Signal	What it actually tells you	What it does not
Domain Rating	Backlink profile strength, useful for search visibility	Editorial quality, legal standing, real readership
Press-council status	Whether an independent body has examined the publisher	Audience size, article quality, whether they will publish you
Traffic estimate	Rough scale of the audience	Audience relevance, or whether traffic is real
Link attribute	Whether a placement passes ranking signals	Anything about credibility. Sponsored is correct practice, not a red flag
Tier or grouping	Reach and publisher type at a glance	Which one is right for your campaign. Regional often beats national

Where to get the data, honestly assessed

Nobody builds a media list from one source. Here's what each source is good for, and where each one lets you down.

Press council registers. Free, authoritative, and the only place you get verification status. Downside: no domains, no contact details, and the tables are usually loaded by JavaScript so you cannot simply scrape them with a basic HTTP request. Plan on either a headless browser or manual paging.

SEO tools. Ahrefs, Semrush, Moz. Fast for authority and traffic at scale, especially via batch lookup. Downside: they know nothing about editorial standing, and batch features usually sit behind a paid tier. Free authority checkers exist but rate-limit hard.

Commercial journalist databases. Strong on people and contact details, weak on Southeast Asian regional press. Coverage of Indonesian provincial media in the big Western tools is thin, and that's exactly where a lot of the useful placement sits.

Your own placement history. Underrated, and the only source that tells you whether an outlet actually published for you. We pulled our own vendor order history for this exercise and it settled arguments no external tool could. Several outlets that looked excellent on paper had never completed a single order, which means they cannot be sold. That's not a data point you can buy.

The outlets themselves. The masthead page tells you the legal entity, the responsible editor, and often the advertising contact. Slow, but it's the ground truth that everything else gets checked against.

The part everyone skips: keeping it alive

A media list is not a document. It's a process that happens to produce a document.

Set a cadence you can actually hold. For a list in the low thousands, a quarterly refresh of authority scores and verification status is realistic, and it takes a few hours if you batch it properly. Write the cadence down where stakeholders can see it, and put the check date next to the data rather than in a changelog nobody opens.

Why bother being public about it? Because the alternative is worse. A list with no dates gets quietly distrusted, and then somebody builds a shadow copy in their own Drive, and now you have two lists that disagree. Seen it happen more than once.

One thing worth adding to the refresh: track removals, not just additions. When 300 outlets drop off a register in six months, the additions are the boring half of the story. The removals are what create risk, because those are the rows that used to be safe.

What to ask a vendor before you buy their list

If you're buying placement through an agency or a marketplace rather than building your own list, four questions separate the serious from the rest.

1. Give me the verification status per outlet, not a general statement. "All our media are registered" is not an answer. Registered where, in which state, checked when.
2. What's the link attribute on each one? Anyone promising dofollow across an entire list of established newsrooms is either mistaken or not reading the outlets' own policies. Many mark paid articles as sponsored, which is correct practice on their part.
3. When was this list last checked? If there's no date, there's no maintenance.
4. Has each outlet actually published for you before? Availability on a rate card and a completed placement are different things.

None of those questions are unreasonable, and none of them require the vendor to disclose pricing or supplier relationships. If a vendor treats them as intrusive, that tells you something.

For what it's worth, we published our own answer to this. The Arfadia media directory lists more than 1,700 outlets in Indonesia and internationally, with press-council status shown per Indonesian outlet, Domain Rating as a number, the check date alongside the data, and sites that accept paid articles but are not news publishers separated into their own labelled tier. No login, no pricing, because rates move per outlet and per format and one published number would be wrong for almost everyone.

A short honest note on limits

Look, no media list is complete. Ours isn't either.

Country was missing for 826 of the international rows in our own dataset when we first built it, because those outlets sit on .com domains and country cannot be inferred from a domain suffix without guessing. Closing that gap took three passes. Two independent audits first, checked against each other on the 370 rows where they overlapped, agreeing on 89.5 percent. Disagreements were adjudicated by source quality: an About page, Contact page, footer or Wikipedia entry beats an aggregator listing. Then every row still resting on an aggregator was reopened and checked against the site's own pages. That third pass moved 68 rows up to a verified source and pushed 34 down to blank, which is the direction that tells you it was done properly. What remains is 104 rows with no country, and blank they stay, because those sites genuinely do not disclose one anywhere. Verification data covers cyber media only, so an outlet verified as print, television or radio carries no mark even though it is verified elsewhere.

Two errors that cross-checking caught are worth repeating, because both would have looked perfectly plausible sitting in a spreadsheet. One audit put telegram.com in the United Arab Emirates, having confused it with the Telegram messaging app at telegram.org. The domain actually belongs to the Worcester Telegram and Gazette in Massachusetts. Another put vanguardia.com in Spain, confusing it with lavanguardia.com in Barcelona. It is a Colombian daily, published in Bucaramanga since 1919. A single source would have shipped both.

The third pass turned up something stranger. Two sites in the list publish addresses that do not exist. One gives a street in "Zynthorian, WV 58204", another "Quarlis Springs, MQ 11223". Neither town is real and MQ is not a US state code. Both rows are blank now. It is worth knowing that a plausible-looking address block in a footer is not evidence of anything until you check the place is on a map.

Those gaps are in the data because being wrong quietly is worse than being incomplete loudly. Any list that appears to have no gaps has simply hidden them.

Frequently Asked Questions

What is the difference between a media list and a journalist database?

A media list is a list of outlets, built around domain, authority, vertical, accepted formats and link attributes. You use it for distribution and paid placement. A journalist database is a list of people, built around names, beats and contact details, and you use it for pitching earned coverage. Both are legitimate but they answer different questions, and one spreadsheet trying to do both usually does neither well.

How many media outlets in Indonesia are actually verified?

Based on a direct pull from the Dewan Pers portal on 2 September 2026, 988 cyber media outlets are recorded, of which 830 hold factual verification and 158 hold administrative verification. Dewan Pers itself reported 809 verified-factual cyber outlets as of May 2026, and 1,277 verified-factual media across all formats. Against an estimated 43,000 online news sites, the verified share is roughly 2.3 percent.

Does a high Domain Rating mean an outlet is credible?

No. Domain Rating measures backlink profile strength and is a proxy for search visibility only. It says nothing about editorial standards or legal standing. Regional newsrooms below DR 30 can hold full press-council verification, while aggregators above DR 50 may have no registration at all. Use authority and verification as separate columns, not as substitutes for each other.

How often should a media list be refreshed?

Quarterly is realistic for a list of a few thousand rows, and it matters more than it sounds. Indonesian press-council certificates run five years, and roughly 300 outlets were removed from the official register between December 2025 and June 2026 for not renewing. A verified-media list exported a year ago is very likely inaccurate now. Publish your cadence internally and keep the check date next to the data.

Can I scrape a press council register to build a media list?

Partly. The Dewan Pers data tables are loaded by JavaScript, so a basic HTTP request returns an empty shell. You need a headless browser or manual paging. More importantly, the register lists media names and legal entity names but no domains, so you still have to map each entry to a website yourself and label how confident that match is.

Should I exclude outlets that are not on the press register?

It depends on the campaign, not on principle. For government tenders, listed companies and regulated sectors, verification usually matters a great deal. For reach inside a single province it often does not, and excluding unregistered outlets would remove working newsrooms with real readers. Since only about 2.3 percent of Indonesian news sites are registered, a blanket exclusion removes almost the entire market.

Are sponsored or nofollow links a sign of a low-quality outlet?

The opposite, usually. Marking paid articles as sponsored is correct editorial practice and reflects an outlet that reads its own policies. A vendor promising dofollow links across an entire list of established newsrooms is either mistaken or not being straight with you. State the link attribute per outlet and filter to what genuinely fits when a campaign depends on link attribution.

What columns should a media list have at minimum?

Six that earn their place, plus one that costs nothing. Verification status and legal entity name answer whether the publisher is real. Domain Rating and vertical answer whether anyone relevant will see it. Link attribute and accepted formats answer what you actually get. The seventh is the date each row was last checked, and a list without it gets quietly distrusted within six months.

---

---