Optimize by Context is the screen where we group your tracked prompts into conversations, one market and one language at a time, and show who wins each one: which brands the AI answers name, which pages they cite, and whether any of those pages are yours. It lives in the Optimize module of the new Opttab, the Agentic Journey OS for brands and agencies. It belongs to Optimization, the third of our four pillars (Discoverability, Visibility, Optimization, Agentic Commerce), and it is where a long prompt list turns into a work plan.
Why we built it
A flat list of tracked prompts treats every question as its own problem. Related questions rarely behave that way. Ask an assistant five versions of "where should I buy this" and the answers tend to lean on the same publishers, retailers and community threads, and to name the same competitors.
That changes where effort pays off. If one competitor page wins across a cluster, one stronger page of yours can answer the whole cluster. If a single publisher is cited across it, one pitch to that publisher reaches every question that draws on it. We built By Context so you can see those clusters without exporting anything to a spreadsheet.
Markets matter too. You track prompts by country and language on every plan, and in a blended view a strong home market can hide a weak one. Here, each market and language pair gets its own list.
Where to find Optimize by Context
Open Optimize in the main navigation. The sub-navigation reads Opportunities, By Context, Site Audit and Agent Pages. Click By Context and you get one screen in four parts: the context list on the left, the selected context in the middle, opportunities on the right, and three page columns further down.
The By Context screen. Contexts are grouped by market and language on the left, the middle card shows the selected buyer question, and the opportunities for that context fill the right, each tagged Content or Outreach with an impact score. The page columns below sort the cited pages by owner.
Step 1: pick a context
The left column groups contexts under market and language headers, for example US English, Canadian French or Dutch for the Netherlands. Each header shows how many contexts sit inside it and collapses, so you can work one market at a time. Every row under a header is a topic, with the number of tracked prompts it holds.
A search box at the top jumps straight to a topic, and the footer shows how many contexts the workspace has in total. We usually start with the market that matters most this quarter, then the topic with the most prompts, because that is where a single fix reaches the most questions.
Step 2: read the selected context
The middle card, labelled Selected Context, shows a real buyer question from that topic, written the way a shopper would ask it. When the topic holds more than one prompt, small arrows page through them.
Below the question sit four rows:
Sources: a row of site favicons, with a count for the ones that don't fit.
Citations: a second favicon row with its own count, for the citations in the answers to this context.
Mentions: the logos of the brands the answers name.
Intent: the buyer intent behind the question, using the same intent stages we use across Prompts, such as Comparison, Recommend or Reputation.
We read Mentions first. If competitor logos are there and yours is missing, the rest of the screen is where you find out why. A site that turns up in both the Sources and Citations rows is one the answers lean on for this question, which makes it worth a closer look in step 4.
Step 3: check the opportunities for this context
The panel on the right, Select Opportunities, lists what we think you can do about this one context. A switch filters it to All, Content or Outreach.
Content cards read "Create content for" a topic, with a reason underneath that tells you how many of the AI answers cite others and not you. Outreach cards point at a specific place the answers already rely on: a community thread to answer, a video or an article whose author you could reach. Each outreach card names the competitor that appears there while you don't, so you know the gap before you click.
Step 4: compare your pages with the pages that win
The three page columns below are the part we use most.
Your Pages lists your own pages cited in this context, with a count. If none are cited, that is the finding. If your site blocks our crawler, the column tells you plainly that it could not fetch your sitemap, and asks you to allow Opttab in your bot protection or add pages by URL with Add Page.
Competitor Pages lists the competitor pages cited here, each with its title, its URL and icons for the AI models that cite it.
Third-Party Pages does the same for publishers, retailers, review sites and communities. It doubles as your outreach list.
With the columns side by side, the question gets practical. What do the winning pages cover, for which intent, that yours don't? Differences in coverage, structure and product detail become a to-do list for whoever owns the page, whether that is the content team, the SEO lead or the person who runs product pages.
From a context to the work
By Context does not change anything on your site by itself. It points your team at the next move, and the rest of the Optimize module carries that move forward.
The Opportunities tab, first in the Optimize sub-navigation, collects the same kind of cards across every context. Filter chips switch between All, Content, Outreach, Discoverability and Agent Readiness. Each group shows its count and a View all link, and each card carries a priority (High or Medium) and, once work starts, a status such as In progress or Draft ready.
Optimize > Opportunities. Content cards state how many answers cite others instead of you, outreach cards name the site to get featured on and why, and every card shows its priority and next action.
Content cards offer Write it for a new piece, Improve it for an existing page, or Open draft once a draft is ready. The writing and review happen in Content Studio, where content gaps also appear under Create Content, together with the domains that were cited instead of you.
Content Studio > Create Content. Each gap is tagged New content or Improve existing, lists the domains cited instead of you, and carries an opportunity score.
Outreach cards offer Draft outreach, or Draft answer for community threads. Once a draft is ready, the card shows Open chat instead, and the conversation it opens also sits under Recents in Opttab Chat, named after the opportunity. These are drafts. A person on your team decides what gets sent.
Behind all of this sits GEO & AEO, our name for turning visibility data into ranked actions: technical fixes such as crawl access, schema, canonicals and product fields, on-page recommendations, content work and third-party outreach. Each action can be tied to the prompts it should lift, with before and after tracking, so you can see whether a change moved the answers it was aimed at. We don't promise a lift. The tracking is how you find out.
Which plans include the optimization work
Technical GEO is on every plan, and so are social posts and backlinks. Guest posts start at Starter. The pricing page has the full plan map.
Getting started
Every context is built from the tracked prompts reported in Answer Engine Analytics. If a market looks thin, the gap may sit upstream: Prompt Coverage shows relevant prompts you aren't tracking yet.
If you already use Opttab, open Optimize > By Context in the account you already use at app.opttab.com. Pick your most important market, find a context where competitor pages are cited and yours are not, and copy the third-party pages from that context onto your outreach list. The launch post on the new Opttab shows how Optimization connects to the rest of the agentic journey.
New to Opttab? Start your agentic journey or Book a demo.
