Sales Call Tracking: A guide to pick your weapon
I recently worked on an interesting problem that my partner’s company was facing- they wanted an effective way to track and record sales calls in order to improve sales training and productivity in the most cost-efficient way possible.
Not having a prior background in sales, I partnered with ChatGPT and looked up to Google to figure out various types of solutions that are currently being adopted in the industry. I was appalled by the plethora of solutions that were being adopted, and tried to break it down one solution at a time. This guide is an effort to put together various types of solutions, so as to serve as a one-stop reference to choose a sales-call tracking solution for your organisation.
TLDR:
Step 1: Identifying requirements
Requirements can vary for each organisation and use-case, and its absolutely imperative to list these down before brainstorming for solutions. I sat down with the sales team to understand what the requirements were, and ended up with a laundry list as below-
· Existing CRM Integration: The company had an existing CRM that they had developed in-house, and the call tracking solution would have to seamlessly integrate with it.
· Easy to scale up and down: As the size of the sales team is extremely elastic, the solution would have to allow for scaling up and down seamlessly, without a lot of cost implications.
· Call timeline tracking: The solution needed to track the complete timeline of a call(inbound and outbound) including creation of the call instance, initiation of the call, ringing, answering and completion of the call.
· Upstream and Downstream recording: Both upstream and downstream audio packets of a call would need to be recorded by the solution, in order to transcribe and/or analyze it in the future.
· Decoupled from mobile internet dependency: Using the solution, call to a lead should be placed irrespective of the sales representative having their mobile internet turned on or off.
· Real-time call streaming: Stream the call in real-time, that can be accessed from a centralized dashboard, in order to stream it to any intelligence services in the future for smart prompting and recommendations during a call.
Step 2: Evaluate pros and cons of different solutions and shortlist the best looking one
As listed in the table above, the types of solutions I found can broadly be categorised into one of the four-
- SaaS-based: A software solution that (mostly) charges per user that can typically be used out-of-the-box.
- PaaS-based: A communication PaaS(CPaaS) solution that provides Voice APIs for placing calls; Requires reasonable development effort, and also charges by the minute for every call.
- Infra-based: Hardware devices that are installed on-premise to interface between PSTN(Public Switched Telephony Network, a.k.a cellular calling) and IP calls
- Android Native: Integration the native Android calling app with a recorder to live-stream call packets to server. This requires root access of the device.
Now let’s dive a little deeper into each type of solution and score it out of 5 based on the following parameters- Ease of integration, Scalability, Cost, Feature flexibility and Maintenance effort. In categories where there are multiple solutions, I have picked one solution to deep dive into.
SaaS-based:
Most solutions under this category had a similar feature-set, that included tracking in-bound and out-bound calls, access to call recording and CRM-integration, which mostly satisfied our requirements. Most of these expected calls to be placed via their mobile app, and some also required active mobile internet to be able to place calls. Some of the solutions explored were TeleCRM, SalesTrail, Runo, CallRail, etc. However, we also wanted to access the live call stream of an ongoing call, and there were only 2 solutions that I found which provided this feature- TeleCMI and TataTele. Also, with these 2 calls could be placed directly via the sales CRM, and they were routed to a lead using the sales representative’s mobile phone.
More on TeleCMI:
TeleCMI itself has 2 types of calling solutions, one where a softphone is used to place calls using the internet, and the other, where the call placed is routed via the agent’s phone to the end user. A virtual number is provided by them for every user, and this is the number that appears as the caller ID to any lead that is dialled. Incoming calls can be configured to stick to the same agent for a given virtual number from their dashboard.
Integration to a custom CRM can be done using webhooks. Call details record and call notifications are sent to these webhooks, based on which actions can be performed.
TeleCMI provides a lot more features like call barging, call transfer, integration with AI services, etc.
Scoring for SaaS-based solutions:
PaaS-based:
C-PaaS based solutions like Amazon Chime SDK, Twilio, and Azure Communication services among many others, provide Voice APIs that can be used to place calls, generate interactive voice responses, record calls and run analytics over call data. However, as these solutions are priced by calls/minute, they work out to be very expensive for any sales-heavy organisation. Addtionally, virtual numbers need to be procured, which come at an extra monthly cost.
More on Twilio:
Twilio offers powerful Voice APIs that can integrated with any CRM to place and record calls, and track other call data.
It even allows one to track the timeline of any outbound call as below-
Twilio has extensive documentation for developers, making it easier to onboard and also providing finer and more precise control to handle call data.
Scoring for PaaS-based solutions:
Infra-based:
This was the solution I explored the least, as scalability was a high priority for the organization, and this solution requires procurement of hardware for scaling up beyond the available capacity. However, it is still a good solution that can be adopted as the cost for setup is one-time only, and maintenance is low once setup is complete. In this type of solution, the hardware device installed contains the SIP trunk, that interfaces between the PSTN and IP network. It needs to be installed on-premise, and usually contains the SIM cards which dial and receive calls. The calls on these SIMs can be forwarded to an agent’s mobile phone using the provider’s dashboard. Pricing for this type of solution is usually the cost of the hardware along with a minimal subscription cost for dashboard access.
More on Yeastar:
Yeastar offers several on-premise solutions, VoIP gateways along with VoIP PBX systems that allows one to connect to traditional lines, VoIP lines, and cellular networks. Inbound and outbound routes can be configured on the trunk on the dashboard provided. In this solution, call recordings and other call detail record data are stored in a hard drive that is plugged into the installed hardware. These files can be accessed via the local network, and can also be accessed via the dashboard.
Scoring for Infra-based solutions:
Android Native:
Integration with a native Android app built in-house can be done using the Telecom and MediaRecorder Android packages to live stream chunks of the call using Voice uplink and downlink streams. The stream can be fed into a multi-modal GPT service to transcribe, summarize and alert incase of any missells. However, the stream from voice call can only be captured by a privileged app, and that would require a rooted phone.
Scoring for Android native solution:
Here’s a quick summary of all the scores-
Ease of integration(5 being easiest to integrate)
Scalability(5 being easiest to scale)
Cost(5 being very reasonably priced)
Feature flexibility(5 having most flexibility of feature addition)
Maintenance effort(5 being lowest effort needed for maintenance)
Based on the above analysis and our requirements, TeleCMI, a SaaS-based solution seemed like the right fit with the lowest TTM and a great set of features that came at a reasonable cost, and hence we went ahead with it.
Step 3: POC the shortlisted solution with a few test users
Integration of our chosen solution, TeleCMI was quite straightforward as documentation was manageable along with good support from their developer team. Documentation for TeleCMI is available here. We integrated webhooks to get call detail record(CDR) and notification data, and also integrated one-click call feature from our CRM that would be routed through the agent’s mobile device.
The CDR from TeleCMI looked something like this-
This also contained the filename of the call recording, which we could then fetch with another API call. The TeleCMI team also provided documentation on how live call feed data(streaming) could be supported from server side.
This PoC was tested with 5 users initially. We used the CDR data to generate detailed reports for every sales representative and also built analytics over this data to get insights into productivity and connectivity rates for every representative. We noted that the connectivity was better than expected and also the data being generated by TeleCMI was quite insightful, and hence decided to scale up the number of users.
Step 4: Scale Up to onboard the entire sales team
The team came up with a gradual scale up model to onboard all sales representatives, to ensure the infrastructure of the organisation is not overloaded at any point of time. I do recommend onboarding an alternative solution as well to ensure that there is no dependence on a single solution alone.
In conclusion, I had a blast trying to solve this problem for the organisation, picking up some great learnings along the way. I highly recommend taking the above-mentioned step-by-step approach if you want to figure out a sales-call tracking solution for your organisation.
In the wild world of sales, picking the right call-tracking solution is like choosing the perfect playlist for a road trip — get it right, and the journey is smooth and enjoyable; get it wrong, and you might end up in awkward silence.
Hope you enjoyed this read, if you know or have come across more solutions, do let me know in the comments below!
