The "modern search": Why modern search fields are actually filters and what that means for users

Has this happened to you recently? You searched for something in Business Central and your search didn't seem to return any results? Even though you were sure that a result should have appeared? Then you may already be using the modern search function!

One of our corporate customers approached us with precisely this problem. To help you avoid facing the same hurdle, this article explains exactly how the familiar and soon-to-be-obsolete (legacy) search differs from the new, modern search. We use specific examples to explain what this means for your search behavior.

While classic (legacy) search fields were genuine text searches that found character strings in databases, modern search fields often function like filter mechanisms. This means that identical entries now deliver different results than they did in the past.

You can find out whether you are already using the trial version of the modern search function by contacting your application administrator. However, the modern search function is not available in all of the search fields within Business Central. Currently, it works in the customer search, for example, but not in the salesperson code search. From BC28 onwards, however, this new search function will be mandatory and will be available in every search bar.

In order to show you the differences between the two search functions, we should first explain the underlying functionality of both variants.  

Legacy search: A classic text search across fields and substring

The so-called legacy search is the original search function in Business Central. It is based on a pure string search, similar to that found in simple search fields.

Specifically, this means for you:

• Each of the terms you enter will be used as search text exactly as it is written.
• Parts of a word (partial strings) can also lead to hits. This means you do not always have to enter the complete term.
• The search is not strictly limited to individual fields. In lists, the search term you have selected is checked across multiple fields.
• Important: The order of the characters you enter is crucial.
  For example, if you enter "Wer," the search will find entries such as "Werner." However, entering "Erw" will not yield the same result, even though the letters are included.

Example 1: Searching for "Who" in the system

In the legacy search, your entry "Who" would have returned the following results:

• Brigitte Werner
• Julia Werner

Both contain the substring "Wer" in the last name.

This search method is therefore very letter- and sequence-oriented and works particularly well if you already know how the term you are looking for begins or how it is spelled exactly.

Modern Search: More of a filter than a search

At first glance, modern search fields look like classic search fields. However, they work differently in the background: technically, they function more like structured filters than a simple text search.

This has the following implications for you as a user:

• The search is usually limited to specific fields. This means that only content from defined fields (e.g., name, number, or description) is taken into account.
• Often, only the beginnings of words are evaluated. This means that your search term must usually match the beginning of a word in order to generate a hit.
• Partial strings often no longer work. If a word contains the search term somewhere in the middle, it will often not be found.
• The order of the search terms entered is flexible. Multiple terms can lead to a hit regardless of their order.

For example, entering "Who Julia" returns the same result as "Julia Who." Both variants find the entry "Julia Werner."
Modern search is therefore less letter-driven and more structure- and field-oriented.

This results in a completely different search behavior for you:
While legacy search checks strictly character by character, modern search interprets the input as a set of filter terms that work independently of each other.

While we focused on an example involving text searches in the previous section, this will not be the only type of search you perform in your daily work. Sometimes you may also be looking for item numbers or numbered projects. However, searching for numbers works on the same principle as searching for text.

Legacy search: Partial string search in numbers

If you now enter the digits "00" in your search bar, you will get the following results:

• 10,000 – Adatum Corporation
• 20,000 – Trey Research
• 30,000 – School of Fine Art
• 40,000 – Alpine Ski House
• 50,000 – Relecloud

All delivered results contain the substring "00".

Advanced Search: Filter by exact numbers or prefixes

Many modern search fields only filter by:

• exact numbers
• or numbers that begin with the entry

If you now enter the two digits "00" in the modern search as before, you will get a different result than with the legacy search. You will not get any hits because no customer number starts with "00." There is also no assigned customer that is solely assigned to the combination "00."

Here you can see similar behavior to that seen previously in the text search: While the outdated search works on a character-by-character basis and finds digits within the number, the filter-based modern search works more stringently.

By now, you should have a good overview of how modern and legacy search differ. In the previous paragraphs, we focused primarily on the general functionality. But what about the order of the strings entered? To illustrate this, let's take another look at the example of searching for Julia Werner's contact details.

The legacy search: A structured sequence

Instead of just entering "Who" as in the basic example explaining legacy search, we now search more specifically for "Who Julia."

Result: The search does not return any hits.
Why?
Because the legacy search looks for the exact string as entered.

However, if you now enter "JuliaWer, "this string will produce a result. The reason for this is clear: this exact string appears in the data record.

Modern Search: A Flexible Order

If you now search for "Wer Julia" in the modern search engine, you will immediately get the desired result for :

Julia Werner
This works because both terms are checked independently of each other.

As a cross-check, you could now enter the logically correct sequence "Julia Wer" again. You will also get the same result, as the order of the partial terms entered is irrelevant in modern searches.

As you can see, the order of the search terms differs significantly in the two variants. The character-oriented search behavior of the legacy search requires you to strictly adhere to the order of terms in the system, whereas this is not necessary in the modern search.

Now we have searched for (partial) names and (partial) numbers. But what happens if we only search for a few characters? In the following example, we are only searching for the two letters "ho."

Legacy search: Field-independent

If you now use the outdated search function to search for "ho, "you will get the following two hits in our database:

• Alpine Ski House
• School of Fine Art

If the search is not field-bound, as in the case of legacy search, it will also find the letter sequence "ho" within a word such as School. The entry does not necessarily have to be found at the beginning or end of the debtor. It can also occur anywhere within the data record.

The modern search: Field-bound

Entering "ho" in the modern search engine yields a different result than in the outdated one. Here, you will only get the debtor Alpine Ski House, as one of the terms contained begins with the letters you are searching for.

You will no longer find the second customer, "School of Fine Art," using this search.

The modern search filters solely on the customer name and no longer searches for partial strings in other fields.

Modern search therefore requires more precise input and, in this case, delivers fewer results than legacy search.

But be honest: do you search for "ho" when you want to get to the "School of Fine Art"?

The modern search function in Business Central may look like a normal search field, but it works like a structured filter in the background. This means that the modern search delivers fast, accurate, and structured results. There are several reasons for this:

• Performance: Partial string searches, where the system has to search through arbitrary character strings, are very computationally intensive and can slow down the search. Filter-based search logic is significantly faster.
• Structure: Modern data models rely on clearly defined fields. A field-based search ensures that results remain consistent and accurate.
• User guidance: Filters help you obtain structured results instead of being overwhelmed by confusing hit lists.
• Scalability: Cloud systems and large amounts of data benefit from indexed filter mechanisms, which work more efficiently than traditional full-text searches.

• Search terms must be entered more precisely
• Substrings often no longer work
• Word order is no longer crucial
• The filter logic of the system is becoming more important
• You need to adjust your search behavior.

Modern search isn't worse—it's just different.
It's faster, more structured, and more scalable, but less tolerant of inaccurate entries.
Understanding the differences helps explain why modern search fields sometimes appear to "find nothing," even though the data is clearly there.

Now you know what to look out for in your next search with the update to BC28, either now or in the near future.