Project settings
A project's Settings page has ten sections, from the brand's details and its answer feed to the scoring model, the outreach identity, the fee policy and the sending window. This article says what each card does and where its full article is, and which settings live in the account instead.
Updated October 19, 2026Markdown
Every project has a Settings page. Its one line at the top says what it is for: where this client's data comes from, how its pages are scored, and the standing answers every email drafted for it follows. Open it to set up a project, connect a feed or Google traffic, move to a newer scoring model, or change how outreach for this client goes out.
The page is a column of cards with a jump list beside it, in the order below. Each card has its own article; this page is the map. Each card saves on its own button. Owners and admins can change everything; the AI engines, Scoring, Fee policy, Reply reminders and Access cards tell a member when a setting needs admin access, and the Outreach identity and Sending window saves need it too.
Project details
Name, brand name, "Industry or niche (optional)" and the "Target markets" chooser. The industry line is context for AI summaries and briefings; the card says it does not affect what is checked. The markets do: the first country on the list is where every prompt is checked. "Save changes" confirms with "Project updated". The brand domain lives on the visibility tracking card below. See Markets and locations.
Visibility tracking
The "Ahrefs Brand Radar" card connects the project to its answer feed. With one-click setup on, "Create & connect report" makes a report for the project and keeps its prompt list in sync, as long as the brand domain is filled in. Otherwise you pick a report with "Browse my reports" or paste its ID or URL, and save. A platform-managed report shows the country your prompts are filed under and a notice when prompts in the report sit under another country, with "Sync prompts now" to re-file them. "Brand domain" has its own save. A data update then runs every morning at 06:00 UTC. See The answer feed and data updates.
AI engines
A grid of the answer engines this project collects on, within your plan's count, with a line per engine saying what your feed runs today. "Save engines" writes your choice; a selection over the count cannot be saved, and engines over the count are paused, never deleted. See AI engines.
Scoring
The "Scoring model" card names the model this project scores on: the model behind the 0 to 100 priority for outreach, the deck's order and what your promotion rules pick up. Its info hint says projects stay on their model until you move them.
- When nothing newer is published, the card reads "This is the current model."
- When a newer model is offered, owners and admins see an "Update to …" button with "a newer model is available". Pressing it opens a confirmation: moving re-scores every opportunity that has not been actioned yet, anything you have already approved, pitched or won keeps the score it was worked on, and it runs in the background. "Update & re-score" starts it; "Not now" closes it. The toast names the new model and ends "Re-scoring runs in the background." The engine room shows the run.
- Members see that the newer model is available and that moving to it needs admin access.
Pages you already approved keep the priority you acted on, which is the review gate's rule: see Statuses and the review gate. The reasons behind any page's priority are on the deck's "Why this score" line and the "Priority score" hint in the detail panel; the card itself never explains the model.
Outreach templates
The "Outreach templating" card is a pointer, not an editor. It shows "Outreach prompt" with how many of its eight sections are set (or "Not set yet. The AI writes from the model email and templates alone.") and "Sequence templates" with a count. "Edit prompt" and "Manage templates" open the editors on the Outreach page. See Templates and AI instructions.
Outreach identity
Who this client's emails come from. Mailboxes are shared across the organization; this card picks which one this client sends from ("No preference: use any active mailbox" is the default), which of its signatures signs the mail ("None: this client's mail goes out unsigned" is allowed), and an optional "Contact number for invoices". With no active mailbox in the organization, the card says to connect one first. "Save identity" confirms. See Templates and AI instructions and Mailboxes.
Send schedule
Two cards. "Sending window" sets when this client's outreach may go out: tick "Set a window for this project", pick days, a start and end time and a timezone, then "Save window". It narrows the mailbox's own window and can never widen it, and the mailbox's daily cap applies either way. The line "Mailboxes currently allow" shows what you are narrowing. A window that never overlaps the mailbox window saves with a warning and the mailbox window is used.
"Reply reminders" sets how many working days a publisher who replied may leave your latest message unanswered before the Inbox marks the conversation "Nudge?" and the deck deals a follow-up. The default is 4 working days, counted Monday to Friday, up to 30; 0 turns it off, and nothing is sent for you. See Sending schedule and limits.
Fee policy
The standing answer when a publisher asks to be paid. Tick "Set a fee policy for this client", then choose "Decline paid placements", "Negotiate, never above a ceiling" or "Accept up to a ceiling". The two ceiling stances take an amount and a currency, "the most this client will pay for one placement", and every stance takes a note of up to 600 characters that the draft uses word for word. It shapes the draft in your inbox reply box; it agrees to nothing, sends nothing and changes nothing about a running sequence. With no policy set, the draft declines politely and says so. "Save fee policy" confirms. See Fees, placements and wins.
Google traffic
The "Google traffic (GA4 + Search Console)" card feeds the report's Traffic & search screen. Step 1 connects Google one of two ways: "Your Google account", read-only access you can revoke from Google, or "Managed integration", where you grant the shown email read access in GA4 and Search Console. Step 2 binds the project's "GA4 property" and "Search Console property" with "Browse granted properties", and sets the comma-separated "Brand query terms" that make a search count as branded. A daily pull keeps it current; the first bind backfills history, and "Sync now" and "Resync history" sit beside "Save" once bound. Two chips at the top show whether each source has written data. See Traffic from Google Analytics 4 and Search Console.
Access and archive
A line points team management to the Members page. Below it, owners and admins get the "Danger zone": "Archive project" hides the project and keeps everything, and "Restore" brings it back from the Projects list, under "Archived"; "Delete permanently" asks you to type the project's name, then "Delete forever" deletes every prompt, opportunity, sequence and import for good. See Organization settings.
Settings that live in the account instead
These are organization-wide, so they sit under Settings in the account, not in a project:
- Members, roles and invitations: open Settings, then Members. See Members and roles.
- Mailboxes and their signatures: open Settings, then Mailboxes. See Mailboxes.
- Agent tokens for your own agent: open Settings, then Agents. See Connect your own agent over MCP.
- Plan, payment and invoices: open Settings, then Billing. See Trial and billing.
- The workspace's name, an Ahrefs token and the workspace's own danger zone: open Settings, then Organization. See Organization settings.
Good to know
- Prompts, competitors and the pipeline are screens of their own beside this page: see Prompts, Competitors and The engine room and pipeline.
- The sending window, the fee policy and reply reminders are one class of decision: set once for the client, applied to every email drafted for it.