Here's some of the questions we typically get asked. Further questions? Just get in touch.
By centralising product development and deploying faster than the rest of the market (typically in just 20% of the time), we're much more nimble than other software out there. Those cost savings are passed onto you, our customers.
Calculate your pricing estimate over on our pricing calculator.
Our standard contract length is just one year, however additional discounts are available for customers who can commit for longer.
We do, please reach out to us for a direct quote via the ‘demo’ form for more details.
Most deployments of Conversation Analytics require only a one-off automated script writing that will send your calls to us via a secure channel.
In most cases this takes less than one days’ worth of work.
We know how annoying it can be when a platform doesn't let you export your data for further manipulation/aggregation with your other business systems. We never lockdown your data and make exporting easy.
Thanks to the way we've built our products we easily integrate with just about any system that you use alongside any manual/human processes that capture data. Get in touch to discuss our tech.
Monitoring for script adherence is one of the simplest use cases for speech analytics as it’s the same key words and phrases that are read out the customer each time. We build what we call an “Event” within the Listen module – the Event is made up of all of the key words and phrases used within the call script.
Our ability to accurately identify if the script was said will be high as we will be feeding in potentially 15 or 20 different phrases to identify the script, not just one or two phrases.
• Identifying missed opportunities for upsell / cross-sell
• Identifying what top performing agents do differently to weaker performers
• Identifying opportunities to improve CX and reduce customer effort (what’s driving repeat demand, complaints, confusion, frustration etc)
• Reducing or handling demand more effectively (what’s driving channel shift, self-serve failures etc)
• Monitoring for script adherence (regulatory, compliance etc)• Automating QA
• Increasing FNOL
• Vulnerable customer identification
• Call handling improvements (did the agent apologise when something has gone wrong, did they show empathy on Claims calls, are they framing the benefits of the products correctly, their use of positive and negative language etc)
Over-time our Conversation Analytics becomes smarter, recognising common words and terms used uniquely by your organisation.
In fact, on the rare occasion that you need to correct an incorrectly transcribed word, the software will learn to transcribe it correctly next time.
We're proud of our conversation analytics - generally more accurate than Google and Amazon. When compared like-for-like, we have scored up to 93% accuracy.
We benchmark ourselves against Googles ASR, however, it’s impossible to know what the transcription accuracy will be on a specific client without doing a test / pilot on your data first.
The accuracy can vary significantly due to a number of factors:
• Whether or not the calls are recorded in mono or stereo
• Landline vs. mobile
• If the customer is in a quiet office vs outside with background noise etc
• Accents
• How compressed the audio is
• Audio format (WAV vs MP3 etc)
Due to the above issues, one call could have 70% accuracy and another 20% accuracy. Clients with the highest levels of accuracy typically have stereo call recordings, use an uncompressed audio format (WAV) and a high bit rate for the audio.
There is no limit to the number of words, phrases or rules that can be included in a query (we call queries Events within the Listen module).
Simple Events may only have 7 or 8 rules, others can have 25+ depending on what we are trying to identify within the conversations.
The platform will transcribe 100% of your call recordings meaning that every spoken word / phrase is indexed and searchable within the platform and the number of matches for terms and phrases will depend on how many times it appears within your conversations.
The Event builder within the Listen module lets you search for contact types and behaviours within your conversations. We typically build Events for: contact reasons (at multiple levels), CX indicators (complaints, escalations, confusion, frustration etc), Effort indicators (mentions of repeat contacts, mentions of long wait times etc), Agent Behaviours (did they apologise, show empathy, take ownership of the customers query) and Insight Events (product insight, competitor insight, pricing insight) – it’s not uncommon for clients to have 200+ Events listening for unique things within their conversations. This will give you a hugely granular level of detail about what’s driving demand and the experience being delivered to customers.
MOJO-CX© was born out of being frustrated and disappointed with platforms like Verint, Callminer, Nexidia etc. whilst TMAC was providing consultation work for contact centres.
These longer-established platforms are typically significantly more expensive, take months to deploy (vs. weeks for our solutions) and require specialist technical resources to get the most out of them. Other platforms are very much designed by analysts for analysts meaning on top of the cost of the software, companies must recruit expensive specialist resources to drive value from them. Where they have speech analytics, it's often lackluster and inaccurate.
Yes, we only act as the data handler and as such you remain in control of your customer's data. Nonetheless we are dedicated to protecting your and your customers' data.
