Just before Christmas Ofcom issued updated guidelines on making calls from your contact centre and the use of automated technology such as predictive diallers. The full document – “Ofcom’s statement of policy on its persistent misuse powers” – is available at the Ofcom web site.

The guidelines are largely a restatement of those released in  2010  with some changes in detail, including  guidance for the use of Interactive Voice Messaging (IVM) with a new type of abandoned call.

However, the statement contains an unexpected change (A4.12 and onwards). Since 2006 the industry has been working on the understanding that abandoned calls below 3% were lawful or acceptable and would not face action; this update states that this is not the case and there is no “safe harbour”. Whilst the full text indicates that a low level of abandoned calls is unlikely to attract Ofcom’s attention it can be read that to be sure of compliance abandoned calls should be avoided completely.

This ambiguity doesn’t do dialler managers any favours; can they dial predictively or not? The guidelines don’t ban predictive dialling but we can imagine the conversation between the dialler manager and the company’s compliance officer:

Ofcom predictive dialler guidelines

What happens if I set my abandoned call rate to zero?

Undoubtedly, if you simply set the abandoned call rate to zero instead of 3% and don’t do anything else your agent productivity will reduce. You’ll need more agents to achieve the same result. We set our mathematicians[i] the task of working out the impact in a typical dialler scenario.

If the abandoned call rate is set at 3%  39 agents are required.


With a 0% abandoned call rate 41 agents are needed.

Our scenario

10,000 records to be dialled in a 12-hour day

Each record is active[ii] in the dialler system for 175 seconds before it’s considered complete

In this example setting the abandoned call rate to zero and doing nothing else increases agent cost by over 5%.[iii]

How can I use the dialler successfully and be compliant?

The time that each record is active in the dialler is a fundamental productivity factor. If we could reduce the average active time by 15 seconds in our scenario then we would only need 38 agents – over 2% reduction in agent costs compared with today’s 3%-abandoned-rate position and 7% less than the 0%-abandoned-rate position.

The burning question is “how can I reduce active time by 15 seconds?” In an Ofcom-compliant environment the time spent dialling a record and not getting an answer is at least 15 seconds, so if you can reduce the number of times you dial without getting an answer then the job is done.

One answer is Precision Dialling.

The days of setting a dialler to keep trying the same number randomly until contact is made are mostly over – it doesn’t please the customer, you risk facing compliance issues and it’s actually more expensive. Precision Dialling gives you the optimum chance of making contact with your customers because you dial:

  • At the right time
  • In the right place
  • With the right message and medium to catch their attention
  • Matching the right agent skills to each call

So there are no more endless re-dials to the same person with the associated risks, wasted time and effort.

Make every call count using Precision Dialling

Precision Dialling is predicated on the idea that you have some knowledge about each customer and then you set your dialler based on what you know. For example, they might have created a quotation on your website or via an aggregator so you know that they’re currently active; or they could be a long-standing, existing customer so you’ll have some history of when it is they call you. You should therefore be able to look at your inbound records to pinpoint the best time to call that person.

If you don’t have much history of the contact – if perhaps you’re cold calling for example, clichéd thinking can work until you start building your own records. For example:

  • call the over 50s in the afternoon
  • if you know someone might have school-age children, don’t bother during school run times
  • if it’s a London postcode try calling later in the evening in case they work late

It’s important to gather intelligence from all areas of your contact centre – inbound and outbound calls, emails, tweets, chats etc. – which might give you valuable customer information.

Use multichannel communication too

Precision dialling reduces your agent costs but it’s not the only option: multichannel contact is another answer. Midland Heart has combined text messaging and Interactive Voice Messaging (IVM) with traditional dialling to implement a rounded solution to customer contact. Customers are sent a text message so that they can respond when it’s convenient for them. If SMS isn’t appropriate automated voice technology gets the message across. The multichannel approach is convenient for the customer, reduces the time that records are active in the system and lessens staff costs. Dialler campaign cost will decrease and business performance is likely to improve if you dial less often. This seems counter-intuitive but if we assume that your campaigns are measured by business outcome such as sales-per-hour-per-agent, or money-collected-per-hour-per-agent, then dialling less often improves campaign performance.

But my case is different; your calculation is too simplistic……

We agree! Every outbound contact centre campaign is different and the probability of your campaign being close to the figures used in the scenario is absolutely negligible.

If you’d like to know more, get in touch on 01483 494 690 or via enquiries@rostrvm.com and we’ll take you through the statistics – it takes about half an hour via an on-line meeting.  We’ll even help you to work out the specific implications in your campaign.

[i] We’ve modelled a typical scenario by applying “Little’s Law” from queueing theory to a predictive dialler process. Wikipedia provides a good starting point to learn more.

[ii] The time that the record is ‘active’ in the dialler is core of the calculation. It is the average for the sum of ALL time that the record is active in the dialler. That is it is the average over the CAMPAIGN of:

The time spent by the dialler to try to contact the customer (even if there is no answer) +

The time spent in conversation (TALK TIME) +

The time spent in call work (PREVIEW TIME & WRAP TIME)

[iii] The number of agents calculated by our model assumes that all agents are active all the time – 60 minutes in the hour. In reality there is ‘shrinkage’; agents take breaks, attend team meetings and so on. The real world number of agents employed will be higher after allowing for shrinkage.