How to check tender portals in twenty minutes each morning

Check your sources once a day, at a fixed time, in a fixed order, and make exactly one of four decisions on every item you open. Twenty minutes is enough for that, with a separate hour once a week for the work that does not reward daily attention.

Almost nobody works this way. The common patterns are skipping three days and paying for it in one catch-up, or opening a portal with no plan and surfacing fifty minutes later having read half a specification you were never going to bid on.

Why a fixed sequence beats browsing

A fixed sequence caps the cost. You know before you start that this takes twenty minutes, which is what lets it survive a bad week. Browsing has no natural end, so it competes with everything else on your desk and loses. It also leaves evidence: when the sequence is written down you can point at the step that did not happen, which is the basis of a useful coverage audit.

Deciding is the expensive part, not looking. Browsing makes you invent the criteria fresh on every item, so you tire and start deferring, while four fixed decisions cost you one choice from a short menu.

When to run the daily pass

Run it after the sources that matter most to you have published, not before.

For the EU that means after 09:00 Central European Time. The Publications Office begins publication of the next issue every working day according to the TED release calendar, and a notice goes live not later than 09:00 on the publication date of the OJ S issue. The S series runs Monday to Friday apart from certain national holidays, so a Monday pass on TED carries three calendar days of activity and a Saturday pass carries nothing new.

US federal work has a different shape. GSA's documentation states that all active notices in SAM.gov are updated daily while archived notices are updated weekly, so an early pass on SAM.gov sees yesterday afternoon's postings and today's wait until tomorrow. The same applies to the UK's Find a Tender service and to national portals generally.

Timing matters because the response clock does not start when you see the notice. Under Directive 2014/24/EU the minimum time limit for receipt of tenders in an open procedure is 35 days from the date the contract notice was sent, or 30 days where the buyer accepts electronic submission (Article 27), and notices must be published no later than five days after they are sent (Article 51(2)). A notice can already be days into its own deadline when it reaches you.

Where the time goes in each pass

The twenty minute daily pass, step by step

Open the queue, not the portals. Two minutes. Start in one place that already holds everything, meaning the folder your alerts land in or a single saved search page. Count the items first, because a queue of eight and a queue of eighty are different jobs.

Triage on the surface facts. Ten minutes. Read the title, the buyer, the submission deadline, the value if stated, and the classification code, which is a CPV code on an EU notice and a NAICS code on a US one. Five fields, about thirty seconds, then decide once and move. If you cannot decide in thirty seconds, the decision is qualify this week.

Do the act now actions. Five minutes. This is the only step where you touch a portal. Register or express interest where the buyer requires it, download the pack into the shared folder, calendar the submission deadline and the clarification cut-off, and name an owner. None of it involves reading.

Log and close. Three minutes. Write the discard rows, add anything genuinely ambiguous to the weekly list, and stop.

The four decisions, and what each commits you to

Act now means something has to happen today: a short deadline, a mandatory expression of interest, or a clarification window closing. The commitment is an owner and a calendar entry before you close the pass.

Qualify this week means it enters the bid/no bid queue with a name and a date attached. It is not a maybe. A maybe is a discard with better manners. When it comes up, run it through your bid/no bid checklist rather than reopening the triage question.

Watch means the notice is not for you now but tells you something: a framework you want to be inside at renewal, or a buyer entering a market you serve. Watch items get a review date, and the weekly pass promotes or deletes them.

Discard means no, with a reason.

How to log a discard so it never costs you twice

An unlogged discard is a decision you will make again, because the same notice comes back as a corrected version, on a second portal, or as next year's re-tender.

Keep one row per discard, wherever you already work.

Use a closed list of reasons so the column can be counted: scope, geography, size, capability, capacity, incumbent, timing. No free text, because free text cannot be totalled.

Two things then become possible. Your incoming queue can be filtered against the reference list, so a logged discard never surfaces twice. And the reason counts feed your next round of alert tuning, because forty geography discards in a quarter is a filter that needs fixing rather than a monitoring problem.

The weekly hour

Block one hour in the same slot each week, ideally not Monday morning when the daily queue is largest.

Live bid deadlines first, fifteen minutes. Every open pursuit gets four numbers said out loud: days to submission, days to the clarification cut-off, owner, next milestone. Clarification deadlines are the ones that pass quietly, because nothing prompts you and the submission date still looks far away.

Adjacent categories, fifteen minutes. Search one classification branch either side of your core, and one buyer type you do not normally serve. This is the counterweight to a well tuned alert set, which by design shows you only what you already asked for.

Award notices, ten minutes. Read what was awarded in your categories this week, who won and at what value, which turns reading award notices into a standing habit.

Watch list, ten minutes. Promote, keep with a new date, or delete. Nothing survives two consecutive reviews undecided.

Source health, ten minutes. For each source you claim to cover, did it produce anything at all this week? A silent source is either a quiet market or a broken feed, and the two look identical until you check. If it is broken, work it as a portal outage.

What not to do in the daily pass

Do not read tender documents. This is the classic trap and it is worth being blunt about. A pack runs to dozens or hundreds of pages, and opening one is a qualification activity that belongs in a scheduled block. Open one at 08:40 and the twenty minutes are gone, usually on an item you were always going to no bid.

Do not retune your alerts mid pass. Noticing that a filter is wrong is a note for the weekly list; tuning while annoyed at one specific notice is how you end up with a filter set nobody can explain.

Do not run the pass twice. Checking again at four in the afternoon returns almost nothing, because your sources have already published for the day. And do not start writing anything, including a note to the buyer.

How to handle volume spikes and quiet weeks

Keep the routine and move the bar, in both directions.

On a spike, raise the threshold for act now and push more items into qualify this week. The timebox does not move. If your queue is regularly over about forty items you have an alert problem rather than a time problem. Some spikes are seasonal: the US federal fiscal year ends on 30 September, and much of Europe empties out in August, so those two queues should not be judged by the same standard.

On a quiet week, run the pass anyway and spend the recovered minutes on the watch list. Resist widening filters for at least a fortnight, because you are usually reacting to a publication calendar rather than to the market.

How the routine changes for a team

For one person the routine is the whole system, so protect it by making act now the only thing you allow yourself in the morning. The discard log matters most here, because your memory is the only other record that exists.

For a team, split by decision rather than by portal. One person runs the pass and everyone else receives its output, which is the opposite of what most teams do. Splitting by portal produces duplicated triage on cross posted notices and an argument about who owns the gaps.

Then be explicit about who sees what. The bid owner gets act now items with the deadline already in the calendar. Whoever holds the commercial view sees anything above your value threshold the day it lands. The delivery lead sees anything that would consume capacity this quarter, because a win you cannot staff is a problem you want early. The account manager sees anything published by an existing customer.

Rotate the pass weekly rather than daily. A week is long enough to learn what normal looks like, which is what lets you notice a source going quiet.

What happens when you are away

The routine has to be runnable by someone who did not design it, and that is the real test of whether it is written down. The checklist below is the handover document.

Name a deputy before you need one and let them run a pass alongside you at least once. Two failure modes account for most absence damage. In the first, nobody runs it and the gap surfaces months later as a missed tender. In the second, someone runs it, makes reasonable decisions and records none of them, so you return to a queue that looks clear and a week you cannot reconstruct. Both are covered in handing over monitoring.

If you come back to a real gap, treat it as a one off backfill with its own timebox rather than absorbing it into the daily pass.

The printable checklist

Daily pass, twenty minutes.

Weekly pass, sixty minutes.

Common questions

What time of day should I check tender portals?

Run it after your main sources have published. For EU work that means after 09:00 Central European Time, because a notice reaches TED no later than 09:00 on the publication date of its Official Journal S issue, and issues run Monday to Friday. For US federal work an early check picks up the previous business day, since SAM.gov's active notices are updated daily rather than continuously.

How long should daily tender monitoring take?

Twenty minutes daily and one hour weekly, for a firm watching a defined set of categories and countries. The daily figure holds only if you open no tender documents and make a single decision on each item. If your queue is routinely over forty items, the fix is narrower alerts rather than a longer timebox.

Should I read the tender documents during the morning check?

No. Reading a document pack is qualification work, not monitoring work, and it is the most reliable way to lose an hour you had budgeted twenty minutes for. In the daily pass you decide only whether a notice deserves a qualification session.

What do I do when a hundred notices arrive in one day?

Keep the same twenty minutes and raise the bar for act now, pushing everything borderline into the qualify this week queue. Do not extend the timebox, because a longer pass on a spike day trains you to skip the next one. A queue repeatedly that large means the filters are too broad.

How do I stop the same irrelevant tender reaching me every week?

Log every discard with the notice reference, the date and a reason from a closed list of seven, then filter your incoming queue against that reference list. The reason counts tell you which filter to change: geography discards mean your regions are wrong, while scope discards usually mean your keywords describe your own service rather than the buyer's problem.