June 18, 2013 at 12:52Segmentation and work with the client base of a web studio / agency
Hello, Habr! Today I would like to continue our cycle of materials with the life of web studios and digital agencies - and talk about working with a client base, as well as specific principles / approaches to its segmentation.
* By the way, this topic is also discussed in detail in our recently launched free special project of 11 video lectures about marketing, sales and customer service in the studio.
Often, talking with representatives of small studios of 10-15 people, I hear: “ We have 3-4 projects in production and a dozen and a half clients for support, I know them all in person, why do I need some kind of segmentation and some metrics for him? ". In general, this is a pretty reasonable question. Why put up an extra garden if all the clients are cursing anyway?
Firstly, your customer base is not only your current customers, but also “outflow” and “rejection” (I’ll write a little more below). With this in mind, the size of the base of one and a half dozen customers immediately turns into several tens / hundreds, or even thousands, of contacts.
Secondly, assessing the prospects of a client should not only be emotionally expert (“ This client is kind of cool, now he orders almost nothing, but we will soon squeeze it ” or “ Well, they’re some kind of small company, a business card site for them done - and okay, let them pay a couple thousand per month for hosting ”), but also rational, based on the values of real KPIs.
Only in this way will we be able to give a 360 degree assessment and avoid the situation when all the company's efforts are devoted to the development of a large federal brand, for which a banner was made a year ago and which is not particularly eager to continue working with us, but a small client who needs a bunch of everything to do on the Internet, misses the attention of client managers, slowly merging with competitors.
Speaking of working with a client base, the agent market usually means a pool of your current, existing clients. In fact, your customer base is much larger and consists of three large blocks, each of which should have its own work strategy
The idea of segmenting something is to highlight a subset of objects (customers) that we know more about than the entire database (at the level of correspondence to particular signs or groups of signs). Thus, having made a filter on our client base based on some indicators, we build work and communication with clients of this segment taking into account the knowledge that they fall under our filter.
For example, by filtering the level of customer satisfaction with our services, you can assign different managers to two segments, one of which can develop relations with a loyal customer well, and the other copes well with “resetting relations”.
Base segmentation can be built on standard parameters that characterize our customers:
Now let's try to give some examples of more non-standard approaches to segmentation of the client base:
Let us give an example of a rather simple approach in ranking our “failure” - the database of customers who contacted us with a request, but did not become current customers.
The purpose of segmentation is to highlight the most promising companies with which to try to resume communication after some time.
I will make a small remark. Unfortunately, more than half of the companies in the market generally do not keep track of incoming calls. Often a company gets into a CRM (or even just an excel file) only after it has become a customer. This is a very serious mistake, and if you do not keep records of incoming calls, you must immediately begin this process, such a database is one of the important assets of the studio’s business.
So, in our table of records of incoming calls, for each client we introduce three additional parameters, each of which we evaluate on a scale from 1 to 5:
Next, for each client, we summarize these three values. We get a certain “renew” index - the attractiveness of renewing contact with customers - an integral indicator by which we can now sort our database (or break it down into large segments). We rank - and we focus most of our efforts on clients with maximum Renew (in our example, with MiddleProm).
Of course, these three indicators are not a paradigm, you can remove some, add some. In the same way - it is not necessary to make a scale of 5 points, you can limit yourself to the “good / bad” switch.
And in continuation of the topic, a few words about the date of the resumption of communication. At what point in time is it better to resume a dialogue with a client with a high renew index? All individually, but you can offer two options:
Probably, almost all readers have heard about the method of product segmentation of the Boston Consulting Group.
All products are divided into four segments depending on the market share of the product and the growth rate of the niche. I will not stop separately, you can read more, for example, here .
Based on this approach, you can make the "adaptation" of the matrix for a business web studio and segmentation of the client base.
On the abscissa axis, instead of the project’s market share, we introduce the indicator “ customer activity ” - as far as the customer is initiative, he offers to do new projects, asks for services, is interested in something new. In general, how much it is "motor".
On the ordinate axis, instead of the market growth rate, we introduce the indicator “ customer capacity ”»- how much we can offer for the client by our specialization. If we have a large brand that has already launched everything that is possible in the digital environment and is not going to develop a special direction, the capacity is low. If a client just entered the market - and in our area you can offer him a bunch more - the capacity is high.
And now we segment our customers in the same way in four variations:
Another simple way to segment is RFM segmentation. We introduce three parameters: Recency (prescription), Frequency (frequency), Monetary (revenue) and for each client we assign values from 1 to 5. Or if the base is small - from 1 to 2. Select segments:
You can read more about RFM segmentation, for example, here .
You can also segment the database by customer satisfaction. Let's look at the easiest way to calculate such a customer satisfaction index - CSI (Customer Satisfaction Index).
In this article, I tried to highlight both a number of general principles and several interesting approaches. Of course, the topic itself is much more extensive than several techniques that I described today. And as always - often the discussion in the comments is even more interesting than the material itself =)
You may be interested in our other materials about the customer service area in studios and agencies:
And last week, together with Megaplan, we launched a large special project, within which you can get free access to a comprehensive course of 11 video lectures (as well as a bunch of materials, presentations, templates and samples) about marketing, sales and customer service in web studios and digital agencies. There is also a separate lecture on today's topic. More details - www.megaplan.ru/digital .
Sincerely, Andrey Terekhov
* By the way, this topic is also discussed in detail in our recently launched free special project of 11 video lectures about marketing, sales and customer service in the studio.
Often, talking with representatives of small studios of 10-15 people, I hear: “ We have 3-4 projects in production and a dozen and a half clients for support, I know them all in person, why do I need some kind of segmentation and some metrics for him? ". In general, this is a pretty reasonable question. Why put up an extra garden if all the clients are cursing anyway?
Firstly, your customer base is not only your current customers, but also “outflow” and “rejection” (I’ll write a little more below). With this in mind, the size of the base of one and a half dozen customers immediately turns into several tens / hundreds, or even thousands, of contacts.
Secondly, assessing the prospects of a client should not only be emotionally expert (“ This client is kind of cool, now he orders almost nothing, but we will soon squeeze it ” or “ Well, they’re some kind of small company, a business card site for them done - and okay, let them pay a couple thousand per month for hosting ”), but also rational, based on the values of real KPIs.
Only in this way will we be able to give a 360 degree assessment and avoid the situation when all the company's efforts are devoted to the development of a large federal brand, for which a banner was made a year ago and which is not particularly eager to continue working with us, but a small client who needs a bunch of everything to do on the Internet, misses the attention of client managers, slowly merging with competitors.
Customer base - not only your current customers
Speaking of working with a client base, the agent market usually means a pool of your current, existing clients. In fact, your customer base is much larger and consists of three large blocks, each of which should have its own work strategy
- Current customers . A pool of your current customers. As a rule, the main focus from the point of view of customer service is precisely on this segment, which is not always the right decision.
- "Failure . " A large pool of customers who contacted you but did not become your customers. As a rule, this is a rather large segment, significantly exceeding the volume of the first block. A special communication strategy should be worked out for this segment - after all, this is also an important asset of your business.
- "Outflow . " Companies that were your customers, but for some reason left. Similar to the previous paragraph, it is important to set up individual business processes for working with this segment of the client base.
Standard techniques for base segmentation
The idea of segmenting something is to highlight a subset of objects (customers) that we know more about than the entire database (at the level of correspondence to particular signs or groups of signs). Thus, having made a filter on our client base based on some indicators, we build work and communication with clients of this segment taking into account the knowledge that they fall under our filter.
For example, by filtering the level of customer satisfaction with our services, you can assign different managers to two segments, one of which can develop relations with a loyal customer well, and the other copes well with “resetting relations”.
Base segmentation can be built on standard parameters that characterize our customers:
- Type of orders. For example, “promo site”, “e-commerce projects”, etc.
- Money / profitability. With what profitability (or total revenue) we work for a particular segment of customers.
- Balance initiation / maintenance. We distinguish two groups of clients — those who are active themselves and always ask for something new, and those who need to be “persuaded” for a long time — it is obvious that the work strategy will vary greatly.
- Frequency of orders. How often does one or another client use our services.
- Prescription of orders. How long ago was the last time we provided some services to the client.
- Outflow indicators, Churn Rate, life cycle sections.
- Branding, image effect, portfolio effect. How important is the client in terms of placing his projects in the portfolio.
- Customer Loyalty, CSI. How loyal and satisfied one or another client is with our work.
- Development potential. How great is the potential (from the point of view of our services) for the development of the client - or is he already "full to capacity."
- and a large number of other indicators that somehow characterize our customers.
Now let's try to give some examples of more non-standard approaches to segmentation of the client base:
Contact renewal attractiveness index
Let us give an example of a rather simple approach in ranking our “failure” - the database of customers who contacted us with a request, but did not become current customers.
The purpose of segmentation is to highlight the most promising companies with which to try to resume communication after some time.
I will make a small remark. Unfortunately, more than half of the companies in the market generally do not keep track of incoming calls. Often a company gets into a CRM (or even just an excel file) only after it has become a customer. This is a very serious mistake, and if you do not keep records of incoming calls, you must immediately begin this process, such a database is one of the important assets of the studio’s business.
So, in our table of records of incoming calls, for each client we introduce three additional parameters, each of which we evaluate on a scale from 1 to 5:
- Brand strength . How attractive the client’s brand is for us, how much it will affect our portfolio, will give us a positive marketing effect. Gazprom is the top five. The incomprehensible Myloprom is a unit.
- The attractiveness of the order . How interesting was the specific order with which the client came to us? It could be a major brand, but with a small task (or was not ready to pay the amount we needed).
- Failure status . How close were we to receiving the order / winning the tender. We almost won, did not reach quite a bit, the client was satisfied with us - five. They performed poorly, made an unsuccessful CP, there was almost no chance - one.
Next, for each client, we summarize these three values. We get a certain “renew” index - the attractiveness of renewing contact with customers - an integral indicator by which we can now sort our database (or break it down into large segments). We rank - and we focus most of our efforts on clients with maximum Renew (in our example, with MiddleProm).
Of course, these three indicators are not a paradigm, you can remove some, add some. In the same way - it is not necessary to make a scale of 5 points, you can limit yourself to the “good / bad” switch.
And in continuation of the topic, a few words about the date of the resumption of communication. At what point in time is it better to resume a dialogue with a client with a high renew index? All individually, but you can offer two options:
- Shortly before the project’s design . We know with what request the client came and we can figure out when the competitor who won the order should launch it. We get in touch a week or two before the planned deadline. Perhaps the current contractor has already "screwed up" and the client is waiting for us in order to "save the harvest."
- A few months after the opening of the project . The vast majority of companies have problems with quality support - and two months after the launch, perhaps the client has accumulated a large number of complaints and discontent - it's time to try to drag such a client to his support and development.
BCG matrix analogue
Probably, almost all readers have heard about the method of product segmentation of the Boston Consulting Group.
All products are divided into four segments depending on the market share of the product and the growth rate of the niche. I will not stop separately, you can read more, for example, here .
Based on this approach, you can make the "adaptation" of the matrix for a business web studio and segmentation of the client base.
On the abscissa axis, instead of the project’s market share, we introduce the indicator “ customer activity ” - as far as the customer is initiative, he offers to do new projects, asks for services, is interested in something new. In general, how much it is "motor".
On the ordinate axis, instead of the market growth rate, we introduce the indicator “ customer capacity ”»- how much we can offer for the client by our specialization. If we have a large brand that has already launched everything that is possible in the digital environment and is not going to develop a special direction, the capacity is low. If a client just entered the market - and in our area you can offer him a bunch more - the capacity is high.
And now we segment our customers in the same way in four variations:
- Difficult children - activity is low, but large capacity. The goal is to rock the client’s activity in order to transfer him to the “stars” segment.
- Stars - and the activity of working with the client is high, and you can still do a lot of things for him. Everything is clear - we are developing and “pumping out”.
- Dairy cows - activity is still high, but the capacity is coming to an end - a little more, and we have nothing to offer - then activity will decrease and the client will turn into a "dog". New niches and service areas must be sought to expand capacity.
- Dogs - and almost nothing can be done, and the client is "tight." Shoot =) Or give it to younger partners.
RFM segmentation
Another simple way to segment is RFM segmentation. We introduce three parameters: Recency (prescription), Frequency (frequency), Monetary (revenue) and for each client we assign values from 1 to 5. Or if the base is small - from 1 to 2. Select segments:
- Customer 5-5-5 - orders often, last ordered recently, the revenue is large - the premium segment with the corresponding appeal.
- Customer 1-5-5. Brought a lot of money, ordered often, but last ordered a very long time. Ahutng, we need to do something and urgently renew relations.
- Customer 5-5-1. I ordered a lot, often, but almost no revenue from it. Maybe him? Well, or to transfer to partners ...
You can read more about RFM segmentation, for example, here .
CSI - customer satisfaction segmentation
You can also segment the database by customer satisfaction. Let's look at the easiest way to calculate such a customer satisfaction index - CSI (Customer Satisfaction Index).
- We take a relevant sample of the customer base;
- We call them (or online, but the quality of communication is obviously worse) and ask one question: “Rate on a 10-point scale how much you are ready to recommend our services to friends”;
- According to the results, we consider the number of ratings for each point. We distinguish the categories of “attract” (the number of high scores in the range of 8-10) and “detract” (low scores of 1-3).
- We calculate the CSI index = attract (8+) / detract (3-); Of course, it should be much higher than one, otherwise you have serious problems.
- We conduct quarterly assessments, look at the dynamics both for the company and for specific managers and clients.
Conclusion
In this article, I tried to highlight both a number of general principles and several interesting approaches. Of course, the topic itself is much more extensive than several techniques that I described today. And as always - often the discussion in the comments is even more interesting than the material itself =)
You may be interested in our other materials about the customer service area in studios and agencies:
And last week, together with Megaplan, we launched a large special project, within which you can get free access to a comprehensive course of 11 video lectures (as well as a bunch of materials, presentations, templates and samples) about marketing, sales and customer service in web studios and digital agencies. There is also a separate lecture on today's topic. More details - www.megaplan.ru/digital .
Sincerely, Andrey Terekhov