◇Opportunity search in practice

It searches publicly accessible sources for tenders and requests that match what you offer, and stores the hits as entries you can work on.

What it is: an aid that saves you clicking through several portals every day.

What it is not:

• Not a complete overview of the market. Anything that is not published is not found.
• Not an assessment of whether a contract suits you. That remains your judgement.
• Not a replacement for your own channels. The best contracts are never put out to tender.

What you can realistically expect: a manageable number of matching hits per week, some of which are worth pursuing. Anyone expecting a hundred hits a day is searching too broadly.

The range of sources keeps changing — several large portals have closed their open interfaces. What is actually being queried at the moment is shown in the overview.

Enlarge

A handful, and the list changes because portals open and close their access.

What decides it:

• Whether a source can be queried openly. Some provide an interface for this, others do not.
• Whether the terms of use allow it. A site that forbids automated reading is not queried — even if it were technically possible.
• Whether the results are usable. A source with a thousand unsuitable hits does not help.

What has dropped out in the past: several large job and tender portals have switched off their access. That is not a fault but a decision by the operators.

What this means for you: do not rely on the search finding everything. It complements your own research; it does not replace it.

Which sources are currently running is shown in the module overview.

The way the people issuing the tenders write — not the way you describe your offer.

The difference: you may offer “process optimisation for medium-sized businesses”. What is put out to tender is “support with introducing an inventory management system”.

How to find the right terms:

1. Look at three tenders you have won. Which words were in them?
2. Ask a customer what they called the task internally.
3. Search with one term and look at the hits — which words come up there?

This makes it hard for yourself:

• Too general. “Consulting” brings everything and nothing.
• Too specific. A product name that only you use finds nothing.
• Only one term. Use five to ten variants.

And the best test: search for a tender you know. Is it found?

Through a fixed list of industries and sectors that are assigned to a hit — without AI, based on the words in the text.

What this means:

• The assignment is traceable. You can see why something ended up in a sector.
• It is reliable in the sense that it does not change from run to run.
• It is coarse. A hit that fits two sectors does not end up in both.

What you do with it:

• Filter by sector if you serve a narrow field.
• Look at hits without an industry too. Not every text contains the words the assignment depends on — these hits are not worse, just not sorted.

The last point is important: anyone who only filters by their own sector loses the hits that could not be classified. At least skim through the rest.

Usually not. Check in this order:

1. Are the search terms too narrow? Use a deliberately broad term and see whether anything comes up at all.
2. Is the period too short? For a niche field, two weeks without hits is normal.
3. Is the region too small? A single district is too little for some fields.
4. Is the source still running? The overview shows whether a run completed or was aborted.

If a broad term finds nothing either, the cause is technical — then it is a case for support.

What you must not conclude from an empty result: that there are no contracts. It means that nothing appeared in the queried sources, with your terms, within the period. Those are three restrictions at once.

And the realistic expectation for a specialised offer: a few hits a month, not a day.

Daily for fields where things move quickly; weekly for everything else.

Why the frequency matters: tenders have deadlines. Anyone who finds one three days before it closes usually cannot bid any more.

What speaks for daily:

• Short deadlines in your field.
• A lot of competition — whoever gets there first has an advantage.

What speaks for weekly:

• Few hits anyway.
• You do not get round to looking every day.

The decisive point is not how often the search runs, but how often you look. A daily search whose results you go through once a month achieves nothing.

What works well: a fixed slot, twice a week, twenty minutes. Look at everything new, decide, move on.

With four questions, in this order — at the first no, you stop.

1. Can we do it? Professionally, with the people we have.
2. Can we meet the deadline? Not the bid deadline — the delivery deadline.
3. Is the bid worth the effort? For some tenders, the application alone takes several days.
4. Do we have a realistic chance? If an incumbent supplier is obviously ahead, that is an answer.

What you can skip: tenders whose requirements are tailored so that only one supplier can meet them. You recognise them by a cluster of very specific conditions.

Time budget: two minutes per hit for this preliminary decision. If you need longer, put the hit on a “take a closer look” pile and work through it once a week.

By entering the deadline as a task straight away when you take the hit over — not as a note.

What you should enter:

• The bid deadline, with the time. Many close at noon, not at midnight.
• A reminder one week before.
• A second one two days before.

Why two: the first is for getting started, the second for getting finished.

What you should also record: where the bid has to be submitted and in what form. Some contracting authorities require particular portals or formats — not something you want to discover on the last day.

The most common mistake: the deadline is in the tender document that was filed away. Nobody sees it there any more. It belongs somewhere that draws attention to itself.

And if you miss the deadline: late submission is generally ruled out in formal procurement procedures.

Take it over, assign it, set the deadline, decide on the next step — four steps, under five minutes.

1. Take it over. The hit becomes an editable entry with a history.
2. Assign it to the company if it already exists. Otherwise create it.
3. Enter the deadline as a task.
4. Record the next step — “request documents”, “ask a question”, “calculate the bid”.

Why step 4 is the most important: a case without a next step gets left lying. The history later shows you what you had planned.

What you should record even if you decline: one line saying why. If the same body puts out a tender again in six months, you will know whether it is worth it this time.

And the history pays off when someone else takes over the case.

That happens when a tender is published on several portals — and that is the normal case.

What is done about it: obvious duplicates are recognised and merged. What cannot be recognised are cases with a different title or a description of different length.

What you can do:

• Look at both. Sometimes one version contains more details than the other.
• Take one over and close the other — with a note that it is a duplicate.
• Note the source through which you actually bid. In procurement procedures, often only one route is permitted.

What you should not do: keep both as separate cases. Then two people work on the same thing, or you submit twice.

And if a tender appears several times with different deadlines: the deadline of the contracting authority applies, not that of the portal.

For the purpose for which it was published — yes. Beyond that, it gets tricky.

What is unproblematic:

• Reading a tender and bidding for it.
• Storing the details for your internal processing.

What you should check:

• Using contact details for advertising. An address given in a tender is there for questions about that tender, not for your mailing list.
• Passing on content. Tender documents can be protected by copyright.
• Creating contact persons as contacts. That is processing of personal data, about which the person has to be informed.

The practical advice: keep the tender case separate from your advertising mailing list. Someone involved in a tender you bid for does not thereby become a recipient of advertising.

By recording why, in a form that can be compared.

Four reasons that it usually comes down to:

1. Price. We were too expensive.
2. Requirements. We lacked a required feature.
3. Incumbent supplier. Someone else had the edge from the start.
4. Bid. We presented it badly.

Why the classification helps: if after twenty cases you see that reason 2 applied twelve times, you know what you need to work on. If reason 1 came up twelve times, you know something else.

Best approach: ask after a rejection. In public procurement you often have a right to be informed of the reasons — with private tenders, a polite question costs nothing.

And the point that brings the most: record why for contracts you have won as well.

For ongoing review: twenty minutes twice a week. For a bid: whatever a bid costs, and that is more than you think.

The realistic breakdown:

• Review: 40 minutes a week.
• Closer examination: 30 to 60 minutes per serious hit.
• Writing the bid: half a day for simple tenders, several days for formal procurement procedures.

Which means: you cannot bid for everything. The preliminary decision is therefore more important than the search.

A useful calculation: if a bid costs one day and you win every fifth, a contract won costs five days of sales effort. Does that fit the contract value?

If not, the answer is not to write faster, but to bid less — and to choose the right ones.