Contact Centre Productivity - More to it than it seems

I have set out in this BLOG article some key advice for productivity and efficiency management inside a contact centre. Most produce lower results than expected and supervisors and managers rarely have the mathematical understanding of the complexities of the environment to enable them to articulate where, or how improvements could be made. Worst of all is the ignorant bliss of the centre sweating its way into mediocrity. This is often due to poor understanding of essentials and subtleties of impacts regarding their actions and correct interpretation of evidence. Inappropriate action is often implemented to amend disappointing results without analysing the detail first, thereby wasting time and precious resources.

So what is the solution for improving contact centre productivity and efficiency? Firstly, a complete understanding of the components of a contact centre is needed, including the individual pluses and minuses of each component, as well as its possibilities and limitations and secondly, the relationship between the components, both mathematically and operationally. Many errors of judgement are entirely due to misinterpretation of results and the placing of blame and subsequent allocation of resources to the wrong solutions.

The components of a call centre can be roughly organised into human, IT, processes, strategy and data.

My first recommendation is ensure the highest quality of agent performance. Human resources must be adequately trained in methods, techniques and character to do their job as best as possible. It is the most dynamic and difficult component. Harnessing the potential takes a lot of skill and spans several areas of management technique and methods. Human resources invariably affect response rate & value, conversation length and contacts per hour. If we focus on the conversation, as long as the strategy is considered optimal, then the quality of the conversation and delivery of the offer is critical to response rate and will obviously impact conversation length. The conversation length directly impacts calls per hour and IT resources to maintain a constant delivery of sufficient contacts for the agents. Finally, the quality of the agent will directly affect the quality of the contact experience the client receives and hence a direct impact on client satisfaction.

Flexibility and power of IT infrastructure is critical to maximising potential. IT resources are put in place to create an efficient and highly productive environment. However it does not always follow that all projects are processed well. Always ensure that as much automation as possible exists within the environment and that it is fast, flexible and easy to use. Everything from dialling, work flow control, fulfilment, confirmations, transactions and reporting are all part of the system. It should also be real-time and integrated with back office functions, data and telephony/web connections. Poor IT infrastructure and capabilities will cause low contact rates, higher costs of processing transactions and lower income per agent hour.

Processes should be in component form and reusable as well as easily connected to do any project. Processes are the methods employed to do projects. It starts with the design of the project, the way in which offers are made, the structure of the work flow (conversation), integration with back office functions and fulfilment, as well as feedback and processes required for data importation and transformation. Processes also include standard forms for call backs, appointments, orders, complaints and much more. Poor processes will affect time per contact, processing of transactions and other back office functions and perhaps even the quality of experience the client will receive at every contact.

Strategy is a major portion of the success or failure of a project. Given good data, great offer, excellent agents, the strategy may still fail. Strategy affects timing, style, positioning, offer, targeting, processes and quality issues. The best advice I can give here is to make strategy the cornerstone for all other components. Be open to altering strategy after reviewing statistical feedback.

Data must be rigorously checked for quality, consistency, accuracy and appropriateness before commencement of any project. Data is often relegated to a secondary importance in most bureaux, as it is seen as client property and as long as it behaves as expected it is not offered much preparation time. This approach is ill advised. Data should be clean, validated, normalised and appropriate for the strategy and processes requiring the data. Duplicate clients should be minimised, telephone numbers and postcodes should be accurate and present, associated table files for data validation of client coding should be present and correct. Nothing is worse than a struggling campaign because of poor data management and applications which cannot manage well without the right data or valid data being used.

Assuming you have great agents, fantastic IT resources, efficient processes and killer strategies, then why are we still experiencing demoralising results or unexplained happenings. Invariably it is a combination of things and we do not always see contact centre management address the most important aspects, but rather attack what seems the most obvious. It’s a bit like the ‘hitting your head against a brick wall’ syndrome. We have all experienced this haven’t we? There is a more analytical and simple way of managing the environment, or being more likely to pin the problem on the most likely cause. It requires application of some simple mathematics and statistical ways to view the results. I have gained some detailed experience over the past decade and a half covering all manner of direct marketing and contact centre analysis techniques and methods. I have summarised for your benefit, a combination of basic data (facts), aggregation, segmentation and interpretation that will be imperative to making the best decisions.

What you must do by looking at the symptoms and evidence, is come to a prognosis about what is the likely cause. From that you can use your knowledge and experience to create a cure. I have tried to simplify the environmental metrics so that you too can understand and improve efficiency or productivity. Firstly, it is important to know what is possible within your environment and therefore you should exercise some caution as to what is truly plausible given your assets and environment. Once possibilities are known, you can set expectations against which you can then measure levels of efficiency.

Another factor that must be considered when setting efficiency and productivity targets is the cost per call or cost per order. Becoming more efficient, doesn’t necessarily imply most productive, as cost to achieve the highest level of efficiency, may mean lower returns. Indeed, every project undertaken by the centre should have an optimal level set that supports maximum return. If you alter anything in your environment or assets, your potential will change and therefore relatively, so will your level of performance. Be sure that you do not continue to measure yourself against an outdated standard, as you will overstate your efficiency relative to your new capacity.

One of the false assumptions about outbound call processing, is that if you have X number of people to call, then you continue dialling as often or when required in order to get all or most completed. This is not efficient or plausible. Indeed, it can be proved that the more you need to try to contact people, the less likely that they will answer the phone. That is, the answer rate drops the more you try. When this occurs, you need to do more dialling per hour in order to maintain a given answer rate volume per hour. This then puts a strain on resources and costs more to process.

Generally speaking the following interpretations can be made about the results from statistical analysis. If the quality of conversion drops below the Required BEP (break-even-point) conversion for the first dial, each subsequent conversion rate from second dial onwards has a marked degree of difficulty, whereas;

If the quality of conversion is above the Required BEP conversion for the first dial, each subsequent conversion rate on second dial onwards becomes easier as the base cost is covered. The Average Call Rate % should be used in the Predictive Dialler for aiding the calculation of the number of required virtual nodes (for outbound dialling). The diminishing profitability per dial stage is directly attributable to answer rates and conversion rates, plus the average income per conversion.

Wait time is affected by risk of abandoned calls dependent upon number of agents, diallers, answer rate and average talk time. As can be seen above, using several average talk times and a variable WAIT Time, one can see the impact on call rates. The percentage spread of average conversation time, is dependent upon several factors, all of which are quality related and not dependent upon a dialler. To optimise a dialler’s efficiency, close management of the following factors should be made:

Suitability & Quality of List
Timing of calls
Call Back Volume (retries) and Timing (set in Dialler Parameters)
Number of Diallers (Routers) Allocated from time to time during the job
Average call length (monitor these by agent and determine any improvements in agent performance)
Number of Agents (the more the better, as algorithms can spread risk more efficiently, optimum performance will start from 10-20 agents)
Answer Rate (how many calls dialled get answered for transfer to free agents)
Abandoned Rate (set at 1percent for EEC compliance) but can be varied

As a rule one can say that ALL CALL RATES related to Calls/Hr are EQUAL in so much as they represent the feasibility of maximum calls per hour (60 minutes) for any given average conversation length and common wait time. Therefore if we are to measure one set of basic expectations against live results, it is important to maintain relativity rather than imposing an impossible set of circumstances which just would not prevail.

The Wait Time between calls can only be reduced by minimising the possibility of abandoned calls, either by increasing the number of agents, or by increasing the abandoned rate (default is 1%), or if there are sufficient agents already and the answer rate has dropped markedly, increasing the number of diallers for the job (routers), will maintain a desired wait time between calls. As can be seen, the variables all contribute to performance and therefore, careful management and analysis of the circumstances should be made before making judgements and affecting any changes in operation.

If anyone would like a set of detailed calculations and models for calculating and predicting efficiency and productivity, please do not hesitate to contact me at clyde.lennon@pd-q.com .