Skip to main content

For publishers and content sites

The evergreen layer under your archive, and the refresh it never gets.

Media sites do not usually need more posts. They need the reference layer their beat implies but nobody has time to build, and a systematic pass over the archive that decayed. Vupie does both, and every draft arrives with its sources attached so your editor can check it in two minutes.

Your archive is the asset nobody maintains

One URL's position, before and after a refresh pass

Decaying Rebuilt
1 5 10 15 Rebuilt and republished Index checked, days 2, 7, 14 Month 1 Month 18 Position

Illustrative shape, not a measurement. The existing URL is rebuilt rather than replaced, so the history and the links stay with it, and the index inspection confirms the update was actually picked up.

Evergreen and reference, not news Archive refresh with index inspection Sources attached to every draft AI provenance metadata on images
Bound periodical volumes and archive boxes on pale wooden library shelving, lit by soft daylight from a window.

You do not need more posts

The reference layer.
And the refresh
the archive never gets.

Your beat implies a set of explainers nobody has had time to write, and your archive holds URLs that ranked once and quietly decayed. Both are work. Neither is news.

Image generated with Nano Banana Pro

Business analysis

What we read before we plan anything.

The 90 day plan is not a keyword list bought from a tool. It is built from what is already connected to your account, which in media means these four things first.

The full method
01

Your archive and what happened to it

Which URLs ranked, which decayed and which never got the update they needed. That list is usually more valuable than any new brief.

02

Your topic taxonomy

So new pieces slot into existing hubs and link correctly, instead of forming an orphaned island.

03

Evergreen queries around your beat

The explainer and reference terms your reporting brushes past but never answers directly.

04

Your editorial standards

Read at onboarding, so what arrives matches your style guide rather than fighting it.

Where the plan aims

The four clusters that carry media traffic.

Priority goes to queries you already sit close to, because those are the ones a well-sourced article can actually move.

Explainers around the beat

"what is [term]", "how does [mechanism] work", "why did [event type] happen"

Reference and glossary hubs

"[field] glossary", "[metric] explained", "[list] by [attribute]"

Decayed archive

Existing URLs that lost position and need facts, structure and links rebuilt

Recurring seasonal coverage

"[event] [year]", "[award] winners", "[season] guide"

The output

What actually lands on your site.

1

Explainer that your reporting can link to

Written once, maintained on schedule, linked from every future story that needs the background.

2

Reference hub

The structured page assistants quote from, with internal links out to your coverage.

3

Archive refresh pass

An existing URL rebuilt rather than replaced, republished, then index-inspected on days 2, 7 and 14 so the update is actually seen.

4

Recurring annual piece

Scheduled to be updated and reindexed before the search season, every year, without anyone remembering to.

Two or three a week

The questions only you can answer.

Research retrieves what is public. It cannot retrieve what only exists inside your business, and guessing it is exactly how AI content gets caught. So before writing, Vupie sends the two or three questions it cannot answer to you by SMS or Slack.

Answering takes a minute, and the answer becomes the part of the article nobody else can publish. In media, they usually look like this.

Vupie SMS or Slack

The archive piece from 2023 outranks the newer one. Rebuild the old URL or redirect it?

Does your style guide require an AI-involvement note on the page, and in what wording?

Which beats are off limits for this? Anything health, finance or legal will be declined anyway.

Answered

The model may cite only from the retrieved source list plus what you answer here. Anything else is refused.

The gate

What we decline in media.

Worth reading before you sign up rather than after. A refused brief frees the slot and you see the reason, but if most of your topics sit on this list then Vupie is the wrong tool for you and we would rather say so now.

What we don't claim

News and breaking coverage

Not what this is. Vupie writes the durable layer; reporting stays with your journalists.

Anything published under a person's byline without their say-so

Attribution is your call, made in the approval step. Nothing goes out wearing a reporter's name by default.

Opinion presented as reporting

Claims are sourced or they are cut, which rules out the comment piece by construction.

Publishing

The usual setup here.

WordPress and Ghost cover most editorial stacks, with categories, tags, authors and feature image set on publish. The webhook feeds custom editorial systems.

Thirteen destinations in total, including a webhook and a hosted blog for sites with no usable CMS. See every integration.

Questions

What media teams ask.

Generated images are written out with IPTC DigitalSourceType provenance metadata, so the provenance travels with the file whatever you decide. How the article itself discloses AI involvement is your editorial call, and the wording is a per-site setting.

Yes, that is the publisher tier. You list your site, set the niches and terms you accept, and every offer arrives with its citation list and quality score for you to approve or decline. It is free, and hosting a link earns credits.

Connect the site. See the plan before anything publishes.

The business analysis runs first and produces a rolling 90 day calendar you can see and change. Nothing goes live until you have looked at it.