Frequently Asked Questions
What Percolated is
-
A tender monitoring service for architecture firms automatically scans relevant tender portals and email inboxes for new opportunities, assesses each one against the firm's criteria, and delivers a filtered report to the bid manager. Unlike generic alert tools, which send raw notices requiring manual review, a managed tender monitoring service scores and recommends opportunities before they reach the team.
Percolated by MACS Bid Studio is a managed AI-powered tender monitoring service built specifically for Australian architecture and design firms. It delivers a daily scored digest — every opportunity assessed as PURSUE, REVIEW, or DECLINE — to the Bid Manager's inbox by 7am each business day.
-
Australian architecture and design firms track tender opportunities through a combination of government portal registrations, email alert subscriptions, and manual daily checks. The main tender portals used by architecture firms include NSW eTendering, AusTender (federal government), Tenderlink, Vendor Panel, and individual local government procurement portals.
Most small and mid-sized architecture firms rely on manual monitoring — a process that misses an estimated 20–30% of relevant opportunities due to alert volume, spam filtering, and inconsistent checking. Percolated replaces manual monitoring with a fully managed AI service that scans portals and inboxes daily, scores each opportunity against the firm's criteria, and delivers a pre-filtered digest every morning.
-
Yes. AI tools can monitor tender portals and email inboxes daily, read and assess each new opportunity against a firm's specific go/no-go criteria, and produce a scored recommendation — all before the team starts their day. This is significantly more effective than manual monitoring or generic alert services, which send raw notices without any assessment or filtering.
Percolated uses AI to score every incoming opportunity against a criteria profile built specifically for each client firm. The result is a daily digest where every opportunity has already been assessed — principals only review the opportunities worth pursuing, not the full volume of what came in overnight.
-
Standard tender alert services — including TenderLink alerts, AusTender notifications, and Vendor Panel digests — send raw tender notices to subscribers. The subscriber must still read every notice and decide whether to pursue it. No scoring, no filtering, no recommendation is provided.
Percolated is a managed service, not an alert tool. Every incoming tender notice is scored against the firm's specific go/no-go criteria before it reaches the team. The client receives a daily digest with clear recommendations — PURSUE, REVIEW, or DECLINE — rather than a raw feed of unassessed notices. It is the only service of its kind in the Australian architecture and design market.
How it works
-
Once a day, between 6 and 7am, ahead of the digest landing in your inbox. If your firm needs more frequent checks, that's available as an add-on at extra cost.
-
A go/no-go scoring system for architecture tenders is a structured framework that evaluates each tender opportunity against predefined criteria before any bid resources are committed. A well-designed go/no-go system considers factors including project type, construction value, client relationship, geographic location, procurement method, competition landscape, and strategic alignment with the firm's business development priorities.
Percolated builds a custom go/no-go criteria profile for each client firm in a 90-minute session with the principal. Every incoming opportunity is then scored against that profile automatically — producing a PURSUE, REVIEW, or DECLINE recommendation with reasoning. Firms using a formal go/no-go process typically reduce wasted bid effort by 30–40% while improving win rates by concentrating resources on the right opportunities.
-
Scoring is only as good as the criteria behind it, which is why calibration is its own dedicated call rather than a form you fill in once. You also get a monthly check-in for the first six months specifically to catch and correct anything that's drifting. If something is ever scored wrong, it's still logged in full in the Audit Log, so it gets caught and the criteria adjusted, rather than the same mistake repeating quietly.
-
The scoring is automated, but the criteria it runs against was set by your firm and calibrated with you directly. Nothing is decided in isolation by AI. Every PURSUE, REVIEW, and WATCH opportunity still goes to your Bid Manager, who makes the actual call.
-
Yes. You get a ready-to-use prompt for summarising opportunity documents with whatever LLM your firm already uses. A long RFT or EOI pack gets condensed to the parts that matter, so nobody has to read the whole thing cover to cover just to decide if it's worth pursuing.
-
If it lands in the tender notification inbox, it gets read and scored, regardless of where it came from. Percolated isn't limited to a fixed list of major portals. It's limited to what reaches that inbox, which is exactly why getting every source consolidated into it during setup matters. The spam and junk folders get checked too, since a legitimate notification landing there is exactly the kind of thing that would otherwise get missed entirely.
-
No. Percolated scores and surfaces opportunities. It doesn't decide what to pursue. That judgement stays with your bid manager, the same as it always has.
Getting set up
-
That's handled as part of setup, not something you need to solve yourself first. Consolidating scattered notifications into one dedicated inbox is covered directly in the discovery and calibration calls.
-
Whoever makes go/no-go calls day to day. Usually a principal, a bid manager, or both. It's a working session, not a formality, so having the right people in the room matters more than the number of people.
-
Setup runs across three calls capped at two hours total, with a test run before you're relying on it for anything real. You're not switching over blind.
-
That's what the check-ins are for. Monthly for the first six months, quarterly after that. Criteria aren't set once at the start and left untouched. They get revisited on a schedule so scoring keeps matching how your firm is actually deciding.
Pricing & contract terms
-
At $350 an hour, manual portal monitoring costs a practice an estimated $45,500 in unbillable principal time a year, before counting what gets missed.
Percolated runs on a flat $500 a month, with a one-off $6,400 setup fee. Paying annually brings that down to $5,100, effectively two months free.
-
A noticeboard shows every listed opportunity and leaves the filtering to you. Percolated scores each one against your firm's own criteria first, so the digest only ever shows PURSUE, REVIEW, and WATCH opportunities, the ones actually worth a look. Anything scored DECLINE or DISCARD stays logged, just out of your way.
-
The system reads the same inbox on the same schedule whether the check-in that month is monthly or quarterly. The price reflects what's running, not how often you're in a meeting about it.
-
Percolated runs on infrastructure MACS Bid Studio manages directly, so that cost never lands on you. The one thing you do need is a Notion workspace. Plus is enough for most firms, roughly AUD $14 to $17 per user a month, and that's where your Audit Log, Go/No-Go Register, and criteria profile live. If your practice runs its own IT security review, Business is worth budgeting for instead, roughly AUD $25 to $33 per user a month, since it adds single sign-on and audit logging that a formal review will usually ask for. Outside of that, everything needed to run Percolated is covered in the setup fee and the monthly fee.
-
The monthly plan has no fixed term, and you can cancel any time. If you've prepaid annually and cancel partway through the year, you get the unused months back. The months you did use get recalculated at the standard monthly rate rather than the discounted one, since the discount was for the full year.
-
No. The monthly fee is flat from day one and doesn't change when check-ins move from monthly to quarterly.
Data & security
-
Just the one dedicated tender notification inbox set up during onboarding. Nothing else in your systems is touched.
-
They sit in your firm's own Notion workspace, visible to whoever you give access to internally. Nothing is shared across clients. Each firm's setup is entirely separate.
-
No. Every scoring call runs through Anthropic's API with their no-training default applied. Your tender content is read once, scored, and never used to improve Anthropic's models or seen by anyone outside your own workspace.
-
Three. Notion holds your Audit Log, Go/No-Go Register, and criteria profile, and it's your workspace, not ours. Anthropic's API does the scoring, under the no-training default above. Make.com runs the automation connecting your inbox to the scoring step and the daily digest. None of the three ever sees more than the one dedicated tender notification inbox you set up.
Proof it works
-
This one's better answered on a call than in writing. Bring a recent real tender and we'll walk through exactly how it would have scored, using an example criteria set close to your own.