Over the past few years, call attribution has quickly become one of the most widely adopted martech solutions. From fortune 50 companies to small businesses, they’ve continually embraced the value of phone calls by attributing millions of offline conversions back to the sources that drive them. Today, marketers are looking to expand how they classify call conversions in hopes of optimizing bidding strategies based on quality customer calls. Below are 4 innovative ways marketers analyze and validate call quality, and the pros and cons to each method.
The most basic method marketers use to assess a call’s quality is by the call’s duration. In practice, calls that reach a specific duration, such as 3 minutes, are deemed qualified enough to be counted as a conversion. This method is used throughout all industries and agencies, but is very common in affiliate marketing as the duration of a call signifies whether or not the affiliate should receive payment for generating the lead.
Pros: Using call duration as a filter gives marketers some control over what qualifies as a conversion. This filter allows the marketer to make educated bid optimization decisions using conversions that were more likely to be actual leads or sales.
Cons: Filtering by call duration is extremely high level and doesn’t offer deep insight into caller intent or true call quality. For marketers that wish to segment call conversions based on sales calls or true conversions (e.g. an appointment booking), we recommend looking at some of the additional options below.
To get a better understanding of the types of calls being driven by campaigns, marketers are beginning to take control of the caller qualification process over the phone. The most apparent change is marketing taking control over the creation and manipulation of IVRs. With the collaboration of sales or operations, marketing departments are beginning to qualify customers over the phone and then route the caller accordingly based on their answers. With this shift in control, marketers can now more easily determine the quantity of calls which are routed to sales or an equivalent department that handles inbound leads.
Pros: By gaining control over the creation of IVRs, marketers now have the ability to easily access data related to the qualification of callers and how they choose to route themselves. With this data, marketers can get a better understanding of whether or not their ads are driving the types of calls they desire. If campaign A is meant to drive new sales, but the majority of calls are being routed to support, then the campaign needs to be optimized or removed.
Cons: Just because someone presses “1” to reach the sales department, does not mean that the call was truly sales related. Callers may press the first number they hear just so they can connect with someone, they could have hit the wrong button, or they could have routed themselves to what they believe was the appropriate department but in fact, it was not.
Keyword spotting is the process of recording and transcribing calls to allow marketers to search for specific words of value. The benefit of this process is that it helps automate the identification of calls where a conversion may have taken place. Many organizations dedicate resources to listening to customer calls in order to identify sales, appointments, and other conversions. The downside is that the employees have to sift through thousands of calls, both valuable and invaluable.
Keyword spotting technology allows users to apply groups of words like“confirmation number”, “appointment time”, “see you on Monday”, etc. as filters to lists of calls with the goal of only identifying calls where a conversion may have occurred. Now the resources that were dedicated to listening to both good and bad calls can now focus on only calls that matter.
Pros: Keyword spotting saves organizations time and money by automating the identification of good calls.There is less cost associated with the process if the company decides to dedicate resources to listening to those calls to identify the outcome. Keyword spotting has other beneficial use cases such as identifying customer interests and determining if callers are receiving a good customer experience.
Cons: The biggest debate to keyword spotting is that it requires organizations to know the words that are being said on the call. For example, if a company wanted to use keyword spotting to automatically identify a negative customer experience, they have to come up with a list of words that customer may use to convey their negative experience (e.g. “not satisfied”, “unbelievable”, “may I speak to your manager”, etc.). However, these specific words may not have been used and the calls were a negative experience occurred may go unidentified. A second con is that many keyword spotting providers use poor quality transcription engines making detecting these words and phrases more difficult. We recommend using keyword spotting applications that rely on the Nuance Transcription Engine (NTE), the most accurate product on the market.
Automatic Call Categorization
The most recent and exciting technology marketers are adopting to analyze call quality is automatic call categorization. Using machine-learning technology and statistical model building, customer calls can be automatically grouped based on various insights marketers are looking to acquire. For example, calls can be segmented into good leads, bad leads, and leads that warrant further analysis.
Automatic call categorization can also be used to improve the customer experience over the phone. If marketing campaigns are generating leads but the customer calls are not being picked up, marketing can provide actionable data to sales and operations to determine if more employees are needed at specific times and days. Not only will this improve customer experience, it will also decrease the chances that potential customers purchase from competitors.
Pros: Call categorization is the most advanced technology available to date for analyzing call quality. It automates the identification of phone leads with an average accuracy rate of 80-89%, thus saving businesses time and money typically used listening to call recordings. Call categorization is extremely scalable and due to its use of machine-learning technology, marketers can automatically group calls based on almost anything.
Cons: In order to obtain accurate results, you need to have a decent volume of customer calls. There need to be sufficient examples for each call category to properly teach the machine-learning algorithms to automatically identify a calls type. If you wish to use call categorization technology but don’t have significant call volume, it is recommended that you use keyword spotting technology until enough examples have been acquired.Business & Finance Articles on Business 2 Community