Focia
Focia is an Australian AI platform that scores content ideas before they go live. It ranks and compares drafts, breaks down engagement triggers and forecasts performance for YouTube, Instagram and TikTok through three platform-specific models.
What is Focia?
Focia is a content prediction tool built by Focia Pty Ltd, a company based in Sydney, Australia. Its premise is narrow and clearly stated: rather than reporting on how a post performed, it scores content before it is published, so that the choice between competing ideas rests on something more than instinct.
Six capabilities are named on the site. Rank & Compare orders and compares ideas by how engaging they are predicted to be. Predictions forecasts engagement metrics across several channels. Content Analysis breaks a piece down into the triggers behind its engagement. Feedback returns written guidance on what works and what needs improving. Workspaces tailors insights and predictions to the data held in each workspace, and Always Learning refers to models the vendor says are retrained in real time on the latest data.
The prediction work is split across three named models, each marked as live. Blaze covers YouTube: it analyses thumbnails and titles, examining thumbnail elements and title hooks and benchmarking them against current trends. Neon covers Instagram, evaluating images and captions against what the vendor describes as an extensive knowledge of the platform's visual aesthetics and engagement patterns. Phantom covers TikTok, having studied large numbers of high-performing short-form videos to judge style, structure and hooks.
Four use cases are put forward: boosting engagement, A/B testing ideas side by side, measuring performance across platforms, and conceptualising ideas when the direction is still open. Two figures accompany them, both attributed to the vendor and neither documented: creators who have integrated Focia into their workflow are said to see an average uplift of 15% in engagement, and the algorithms are claimed to predict content performance on multiple platforms with over 80% accuracy. The site's own metadata mentions e-commerce content alongside social content.
The product itself is a web application hosted at app.focia.io. The public site, by contrast, is a single page. It carries no pricing, no sign-up, no documentation and no legal page; the tariff, product, blog and legal pages that were online in 2023 have since been removed. The only call to action, repeated three times, is a contact form.
What it does
- Rank and compare several content ideas by how engaging they are predicted to be
- Forecast engagement metrics across several channels before anything is published
- Break a piece of content down into the triggers that drive its engagement
- Return written feedback on what works and what needs improving
- Score YouTube thumbnails and title hooks against current trends, through the Blaze model
- Assess Instagram images and captions against platform engagement patterns, through the Neon model
- Evaluate the style, structure and hooks of short-form TikTok videos, through the Phantom model
When to use Focia / When not to
A quick filter to help you decide if Focia is the right fit.
When to use Focia
- YouTube, Instagram and TikTok creators who want a second opinion on a thumbnail, a title or a hook before publishing
- Social media managers who have to pick one post out of several competing drafts
- Growth and performance marketers who track content performance across more than one channel
- Designers, video editors and copywriters producing thumbnails, captions and short-form video hooks
- Founders and small teams running their own social presence without an in-house analytics function
When not to use Focia
- Buyers who need published pricing before opening a conversation, since no plan or rate appears anywhere
- Organisations with procurement or legal review, as the site publishes no terms, privacy policy or data processing agreement
- Developers looking for an API, an SDK or any technical documentation
- Anyone who needs a mobile app, as Focia has no iOS or Android release
- Teams publishing mainly on LinkedIn, X, newsletters or blogs, which none of the three models covers
How to use Focia
A typical end-to-end flow, from setup to results.
- Open the single-page site and watch the embedded product video, which covers the whole pitch in one scroll
- Read the three sections the navigation exposes — Home, How It Works and Models — to see which platform each model handles
- Use the Contact Us button, the site's only entry point, which opens an external form
- Submit the form and wait for the team to reply; no self-serve sign-up is documented anywhere on the site
- Once granted access, open the web application at app.focia.io in a browser, as it is rendered entirely in JavaScript
- Submit the content ideas you are hesitating between, so they can be ranked and compared against each other
- Read the content analysis, which breaks each idea down into the triggers behind its engagement
- Use the written feedback to see what works and what needs improving before you commit to a version
- Check the predicted engagement metrics for the target channel, produced by Blaze, Neon or Phantom
- Keep working inside the same workspace, which the vendor says tailors insights and predictions to its own data
Pros & Cons
Pros
- Acts before publication rather than reporting after the fact, which is where the decision is actually made
- One dedicated model per platform instead of a single generic scorer
- Side-by-side comparison turns a gut call between drafts into an explicit A/B test
- Looks at the elements that actually carry engagement: thumbnail, title, image, caption, video hook
- Workspace-level personalisation is advertised, so scores are meant to reflect your own audience
- Models are described as retrained in real time, which matters on platforms where trends turn over weekly
- Runs as a web application, with nothing to install and no code to write
Cons
- No pricing is published at all — no plan, no entry price, no billing unit, no currency
- No self-serve sign-up: the only route to the product is a contact form and a wait
- No terms of service, privacy policy, data processing agreement or security page; all existed in 2023 and have been removed
- The two headline figures — 15% average uplift, over 80% accuracy — come with no methodology and no source
- Coverage stops at three platforms, and no third-party integration is announced
- No API, no mobile app and no documentation of any kind
- The public site is down to a single page, which leaves very little to assess before making contact
Pricing & Plans
Focia publishes no pricing information. No rate card, entry-level price, billing unit or currency appears anywhere on the site, and the pricing page that was online in 2023 now returns an error. Neither a permanent free plan nor a free trial is mentioned. Access is arranged commercially: the only route offered is a contact form, so the cost of the tool has to be obtained directly from the vendor.
Data, GDPR & hosting
A consolidated view of how Focia handles your data.
GDPR overview
There is no mention of the GDPR anywhere on the Focia site. No privacy policy is published, no data protection officer is named, no Article 27 representative is designated, and no email address is offered for a data subject request — the only channel is an external contact form. Focia Pty Ltd is established in Sydney, outside the European Union, and nothing indicates a European establishment or a transfer mechanism. One European reference does appear, in the site's robots.txt: a content-signals block reserving rights under Article 4 of EU Directive 2019/790. That text concerns copyright and text-and-data-mining, not personal data, and must not be read as any form of GDPR compliance. A privacy policy did exist on the site in 2023 and has since been removed.
Data retention & training
Things to keep in mind
Risks and trade-offs to weigh before adopting Focia.
- Letting a predicted score stand in for editorial judgement, so the model quietly decides what gets published
- Optimising every post on the same engagement signals, which flattens a distinctive voice into whatever currently performs
- Treating the advertised 15% uplift and 80%+ accuracy as measured facts; neither comes with a published methodology
- Confusing predicted engagement with quality — tuning a hook can push work towards the sensational without anyone deciding to
- Uploading drafts, images and unreleased ideas to a service that publishes no terms of service and no privacy policy
- The site says models are retrained in real time and that workspaces tailor predictions to their own data, without ever saying whether customer content feeds that training
- Making first contact through a third-party form, so the details entered pass through a processor the site never names
Setup & Integrations
Technical difficulty
Technically there is nothing to set up: Focia runs as a web application, requires no installation, no code and no API key, and the only prerequisite is a browser able to run JavaScript. The real friction is access rather than configuration. There is no self-serve sign-up, so getting in means submitting a contact form and waiting for a reply. No documentation, tutorial or help centre is published, so the learning curve inside the product cannot be assessed from the outside.
Deployment
Behind Focia
Fundraising
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What does Focia actually do?
Which platforms does it cover?
What exactly does it analyse?
How accurate is Focia?
How much does Focia cost?
Is there a free plan or a free trial?
How do I get access?
Is there an API or a mobile app?
What does the site say about privacy and the GDPR?
Who is behind Focia?
Should you pick Focia?
Focia addresses a real and well-defined moment in a creator's workflow: the point where several ideas are on the table and only one will be published. Scoring thumbnails, titles, captions and video hooks before release, with a separate model per platform rather than one generic scorer, is a defensible design, and the workspace-level personalisation the vendor advertises would matter to anyone whose audience does not behave like the average.
What the site does not provide is the material to judge any of it. The headline claims — a 15% average uplift in engagement, better than 80% prediction accuracy — carry no methodology, no sample and no source, and should be read as vendor statements rather than measurements. Nothing published shows what the interface returns, how a score is explained, or how it holds up when a trend turns over.
The commercial and legal picture is thinner still. No price of any kind is listed, there is no self-serve sign-up, and access runs entirely through a contact form. More significantly, the site publishes no terms of service, no privacy policy, no data processing agreement and no security page — all of which were online in 2023 and have since disappeared, along with the pricing, product and blog pages. For an individual creator that may be a tolerable unknown. For an organisation uploading unreleased campaign material, it is a gap that has to be closed in writing first.
Worth a conversation if you publish regularly on YouTube, Instagram or TikTok and want a second opinion before hitting publish. Treat that first exchange as due diligence: ask for the pricing, for the contractual terms, and for the evidence behind the accuracy figures.
- Choosing a selection results in a full page refresh.
- Opens in a new window.