Segments 2.0 is the new version of Warmly’s lead search engine. It’s built to be simpler, faster and more powerful than the previous version. Let’s explore this new tool so you can unlock its full power and find leads you didn’t even know existed!

Categories & Filters

Segments are now split into five different categories (highlighted in blue) - each covering filters for either company or contact (and sometimes both). This is important so you can build more specific filters and also helps us query data more efficiently.

Operators

This is one of the main changes from the previous version as we tried to make the segment building experience simpler and more uniform.

The Bread-and-Butter

There are two operators which can be found in the majority of fields: Is any of and Is none of. They allow you to build a list of values to either include or remove them from the results, for example:

will get you companies from either of these two industries. On the other hand, if you change the operator to Is none of you will get companies from any industry except those two.

Specialized Operators

Non-text values cannot be filtered using the same logic, but those can be handled using specific operators:

Numeric filters like Opportunity amount let you use Equals to, Is greater than or equals to, Is less than or equals to, Is between, then specify a value or a range of values which the deal amount must be in, the bellow example gets deals of $10K+.

Date filters like Opportunity close date let you use Is, Is after or Is before a specific date, e.g.:

Starts With / Does Not Start With

These operators are currently available on a few text-based filters, but they function with a somewhat nuanced behavior:

- Page Visit Path and Website Form Fill

This will filter the selected values and include whatever text follows up each one, for example:

will get you people/companies who have visited any page under /blog - including itself. So /blog/clearbit-competitors and /blog/warm-leads will also match.

Keep in mind that the selected values are the starting point so /p/blog/any-other-path won’t match in this case.

The Does not start with works similarly but with the opposite effect, for example if you want to exclude people who have only visited any blog posts instead.

- UTM

By default the UTM parameter filter will show you a list of parameter_name=value to select and filter using the already classic Is any of and Is none of operators. But you may want to check for the presence or absence of a specific UTM parameter name regardless of its value. In this case you can change the operator to either Starts with or Does not start with to turn the values list into a parameters only, e.g.:

The above example gets you website visits that included any utm_source.

- The Mighty “Exclusive Filter”

Since CRM data is organized with all its elements related to each other, CRM segments come with one extra option to change how data is queried. Known off the record as “exclusive filter”, this option lets you change how the segment will handle the cardinality of a field. You will see this as the dropdown bellow the operator and search input:

To better understand the purpose of this option lets see how CRM data is organized:

Contacts and deals exists linked to an account, and they are many. Depending on what you're looking for you may set up the filters differently. For example:

“I want to filter accounts which have never closed a deal successfully”

The first thing that may come to mind is to set up is an opportunity stage filter with either is any of Closed Lost or is none of Closed Won. Now let’s consider the following account:

Both filters would match this, but the intended result wouldn’t be right because despite the past closed lost deals, there is a closed won.

In order to narrow down the results and say that you want exclusively Closed Lost deals you need to change the option from At least 1 opportunity to All opportunities. This will automatically remove values other than the ones you selected. Just like that.

Some other examples where the cardinality can make the difference:

• You want companies with all deals owned by a single person - All opportunities should be selected

• Same for contacts or leads owned by a single person

• Not as common as the previous, but you may have multiple branches of the same company under the same domain, e.g. adobe.com, adobe.com/jp, adobe.com/uk. These could be 3 different accounts in the CRM each with its own lifecycle stage and owner, so any filter must include All accounts to make sure the one who matches will not be affected by the others.

The Existence Paradox

All filters are contained inside a higher level data group - the category. Any segment filter added will assume that the data exists in this group in the first place. Consider the following:

When you add a CRM account filter, there must be a CRM account there otherwise it won’t match. This may sound obvious, but if you are using the Is none of operator this may lead to some confusion.

To exclude values from the result a CRM account still must exist. A filter like CRM account stage is none of Customer for example, will return any other account in a different stage, and you may think companies there are not in the CRM will also match since they don’t have the stage Customer to begin with. But that is not what happens since they don’t even exist in the CRM at all. How do you include those then?

To handle this case you must combine that filter with OR CRM account does not exist to tell the segment engine to explicitly include them in the results. The end segment will look like this:

The segment will be executed as the following:

The same goes for Target Customer Profile as it won’t consider unidentified leads whether you are including or excluding values from the results. For that a similar approach is available by adding the OR Company does not exist filter.

For example, if you are looking for a few specific industries, but if it doesn’t match then you want unidentified people (so you can engage with an Inbound Chat). The segment will look like:

Combinators

Now that you have a filter let’s build another one and combine them. Like the classic segments there are 2 combinators available: AND and OR. They may be used in 2 different levels to enable more complex segments. Before we move forward, let’s have a quick “boolean logic 101”:

Each filter will be evaluated individually and will return whether it matches or not. Is the Company Industry any of Internet Services? Are the CRM deals not Closed Won? Then the filter group will combine them and push the result to the upper level, which will do the same with all group results to ultimately get the segment result. Using the AND combinator will require all filters to match so the group end result will be true whereas OR only requires one of them to match.

Segment Level Combinators

Adding the first filter will give you two buttons to add the second filter (and the following ones, as many as needed), this is when you select the combinator for that group:

You can change that later without having to reset the existing filters by changing the dropdown at the top of the group container:

You can add a new group and that will enable the Segment level combinator, which works similarly:

Each group has its own combinator and each of them can be set individually.

Search Level Combinators

Once you build the segments of each category and named them meaningfully you can start using them combined. Will you find the Segment Selector across many Warmly features:

Any segment selected inside a category will be combined with an OR, and segments of different categories will be combined with an AND. For example:

Returns matches for the segments:

• General ICP OR Internet Companies from the Target Customer Profiles

AND

• LinkedIn UTM Source Visitors from Signals/Website Engagement

Operators + Combinators: Warm Leads lie in the details

Combining both is straightforward most of the time and routine, since most of segments will not be made of single filters so often. But there are some gotchas you must be aware of to prevent wrong or inconsistent results. Let’s start with the type of segment you're creating:

Include-Like Filters:

This is probably the most common usage of a segment’s filter, by selecting the Is any of operator (or any other that doesn’t include a negative term) you are telling the segment to explicitly add a certain range of values to your results. If you combine it with another filter of the same category to narrow down the results, AND will be your combinator to go:

Which can be combined with another filter group on the same segment to broaden the search:

This is not a rule of thumb, but filter groups are more commonly used to specify the search even further what you want to see, whereas adding more filter groups will help cover different situations on the same segment.

But if you are not trying to narrow down a search and just cover more options, then the operator must be OR, following the semantics of the Is any of operator as: it can be this, or that, etc

Exclude-Like filters:

Removing values from the results is also another way of using a segment - and sometimes easier, but they are quite more nuanced than they look. If we go back to the Existence Paradox, you’ll see that every filter is contained inside a data group - the category, e.g. all CRM Deals fields are connected by the same group:

Combining filters of two different deal fields will make the segment reference the same deal again. As we've seen before, using the Is any of with AND will just narrow down the search to a more specific result - and that is fine.

On the other hand when you remove a value that you don’t want to see, you are implicitly leaving all the others - which is not always clear:

Specially for CRM data segments - though this example is simplified - real life deals often have a long list of stages, owners and statuses. Combining this filter with another one may lead to a conflict - for example some deal stages exist only within a specific status and pipeline, and depending on how you set up the filters this will get no results.

A deal stage or status may reference an account’s lifecycle stage or status as well, thus making it more prone to confusion. Consider the following segment (which by the way was taken from a real segment):

This segment returns no results since excluding values combined by OR in different objects (account and opportunity) are including a lot of other values to the filter under the hood. If we invert the above filter to rather include what we want to see the list will be massive, so choosing to remove values is certainly the right thing to do. So how should this filter be re-built?

Depending on how your CRM data is organized and which statuses/stages you are using, just changing the combinator to AND could do the trick. Or you can split them by object first and combine both the filters and the filter groups.

Looks like they're doing the same thing, isn’t it? In fact they are, but splitting also make it easier to spot what’s going on when you open the segment definition. It also illustrates how it might be tempting to use OR when combining two different CRM providers, but still AND must be used since the data exists in both.

Again - not a rule of thumb - but when excluding values from the results you'll find yourself using AND more often than you think :)

This article was written by the Customer Success team at Warmly. Please feel free to reach out to your CSM directly or [email protected]