The use of speech analytics in business

    Purchase conversion is a key business metric. Conversion depends on each stage of the sales funnel, from marketing, until the transfer of funds to the account. If the negotiation phase of the sales funnel works inefficiently, it proportionally reduces the overall sales conversion.
    Speech analytics technologies improve conversion rates at the sales stage. The seller’s speech skills directly affect the likelihood of a transaction.

    Working on the development of speech analytics solutions in the DATA4 team, we were convinced that the conversion of vendors who own the negotiation technique and untrained employees differs by a factor of 2-3.

    Using speech analytics technology identifies the patterns of successful technicians, and controls their use.

    Tasks that speech analyst solves:

    1. Control of following the script

    The “script” can be understood as direct reproduction of key phrases and following the dialog box, or mentioning words or controlling the achievement of semantic points (did the operator take the client’s contacts, arrange a visit, etc.).

    2. Identify conflicting contacts

    Detection of conflict situations minimizes outflow, negative comments and promptly signals problems that have arisen. To identify conflicting dialogues, both acoustic analysis for the presence of emotions and lexical analysis by words and word combinations are used.

    3. Determination of the best techniques

    Does the operator work out an objection, which words and expressions does it use? Are there pauses and interruptions in the dialogue? Speech analytics reveal hidden patterns in successful sales.

    4. Identify bottlenecks

    If 20% of calls end with the phrase “there is no product in stock,” or “all managers are busy,” this is a 20% reduction in revenue. With selective listening, the problem may not be identified, but a statistical analysis of 100% of calls reveals even unobvious sales bottlenecks.

    Limits of applicability

    Speech analytics capital cost technology, it is advisable to apply in the following cases:

    1. Contact center more than 10 –15 operators.
    2. The company sells either high value-added goods and services (real estate, cars, money management), or the contact center consists of several hundred operators and automation is necessary to reduce maintenance costs.
    3. Traditional methods of control (listening by supervisors, coaching operators) have already been implemented.

    Technologies used in speech analytics solutions

    1. Speech Recognition to Text (STT).
    2. Detection of speech.
    3. Recognition of emotions by voice.
    4. Lexical text analysis.

    Speech analytics solutions use machine learning technologies. The experience of speech analytics solutions in the DATA4 team shows that it takes 3-4 months to get a good implementation, which includes setting up speech recognition and lexical analysis for a specific case.


    Speech analytics technologies increase sales conversions in the telephone channel. It is possible to increase sales by tens of percent, but there are limitations! The company should already have implemented procedures for traditional control of the contact center. If sellers do not call back to the customer, speech analytics will not help much. The economic effect is achieved either when used in large contact centers, or when selling high value-added goods.

    Also popular now: