Beyond the Hype

Beyond the Hype: Managing Expectations

by Annica Towliat, Senior Product Research Manager, Sony Ericsson
and Jürgen Warnecke, SKIM analytical

This article describes how marketing intelligence allows Sony Ericsson to make a more valid assessment of the capitalised value of new products and services enabled by 2.5 and 3rd generation technology. As standard (conjoint) methodologies do not suffice, a new trade off technique is presented. The article focuses on the practical and methodological problems that made the new technique necessary. In addition, the article describes the types of results and how Sony Ericsson uses them. The article is based on the learnings gained in two global surveys in 2001.

Introduction
The case study
How to assess the value of the mobile phone product line?
How to assess the value of mobile services?
Conclusions
Epilogue
Literature
The authors

Introduction

GPRS and especially 3rd Generation technology have for long been predicted as the developments that will offer the telecom industry vast new opportunities. Therefore, most of the operators as well as the manufacturers of mobile devices and mobile networks have invested heavily in R&D, mergers, acquisitions and in UMTS licenses in order to secure a leading position in the next generation mobile communication promised land. Most of these investments were made, or initiated, at the end of the last decade when the opportunities (and cash flows) in the telecom industry seemed limitless.

Since then a number of undesirable developments have occurred. Markets in the forefront of mobile communication have saturated rapidly, causing a sharp decline in growth. This in turn caused increased levels of competition, thereby shrinking margins. Furthermore, the stock market and especially the technology and telecom stocks have plummeted in value since March 2000. As a result, the opportunities offered by the new technology, and the cash flows that it will enable are considered key in getting the industry back on course again. In other words, the stakes are much higher now.

However, at this stage the value of the new products and services that will be enabled by the technology is uncertain. In retrospect, the introduction of SMS has been very successful and has proved that the market was interested in other forms of communication besides voice calls. On the other hand, the much-anticipated introduction of WAP has been far less satisfactory and is considered by many a flop. These setbacks that have occurred in the last 2 years, coupled with the delayed introduction of 3rd generation technology, have caused many in the industry to doubt the potential of this new technology and the products and services that will be derived from it. It would appear that the original unbridled optimism has been replaced with pessimism.

Sony Ericsson is a key player in the development of new mobile communication products and service concepts. Now that the industry is on the brink of the 3rd generation era, failure is no option. It is vital for Sony Ericsson's product planning and portfolio management to gain insight into the market potential of new features and services such as MMS, online games, video telephony, location based services and downloadable multi-media. Integration of consumer insights in the business model is vital. Understanding the drivers and barriers behind consumer perception and behaviour is powerful and critical fuel for product development. This insight is needed to develop and market a product line that consists of mobile devices equipped with the most attractive features and functions, at the right prices.

The case study

In 2001 Sony Ericsson commissioned two global studies, aimed at establishing the capitalised value of the new products and services enabled by the new technology. To make the expectations manageable Sony Ericsson deemed it necessary to know and understand which concepts generated enough willingness to spend among consumers for the company to be able to pursue the development and introduction of these products and services.

The scope and aim of the studies
Even though Sony Ericsson is known and esteemed as a manufacturer and marketeer of mobile devices, the scope of the issue it faces is not limited to devices alone. Many of the functions that will be enabled by new features on mobile phones are useless to the consumer if the mobile service providers do not support the functions. In addition the fee and / or tariff charged by the operator for using the function will play a vital role in the adoption of the services and features. For instance, an integrated MP3 player allows the user to listen to music and to swap songs via the network with friends. If the operator charges an excessive amount for the downloading and transfer of songs, it is likely there will be less interest in this mobile phone concept.

For many reasons (logistics, interview length, budget, etc.), it was decided to split the task and conduct two studies. By interviewing representatives of the same consumer segments in the same markets the results could be integrated and provide an overall understanding.

The aim of the first study was:

To provide insights in the relative importance of different characteristics when choosing and buying mobile phones.

The aim of the second study was:

To create a better understanding of the preferences and willingness to pay for the use of new services offered in the network.

Preconditions regarding the studies
The main problem faced in the surveys was that we were dealing, in most cases, with completely new features and services for which a frame of reference and a pricing framework was lacking. Extrapolating past behaviour was therefore not a feasible option.

As there was no crystal ball available, two solutions were adopted:

1. Only consumers that are most likely to adopt the new products and services first were interviewed.

2. These consumers were ‘educated' regarding the new products and services by means of carefully developed stimulus material of the highest quality.

Ad 1.
Sony Ericsson's consumer segmentation model was applied. This model provides a screening method to identify consumers that can be considered as true early adopters when it comes to mobile phones and mobile services. In addition, only consumers from the most advanced markets were interviewed: Italy (Milan), Sweden (Stockholm), Taiwan (Taipei), China (Hong Kong) and USA (Boston / Los Angeles).

Ad 2.
A set of show cards was developed that explained the features and services by means of graphics and text. In addition to describing the features and services they were clarified by means of practical examples of the use of these features.

Example of show card used to explain MMS


In total 2000 Computer Assisted Personal Interviews (CAPI) were administered. Per city, 400 consumers were interviewed (200 regarding the sensitivity towards mobile phone features, 200 regarding the value of mobile services).

How to assess the value of the mobile phone product line?

Analysing actual sales figures and comparing that to competitors' sales figures can best determine the attractiveness of a product or portfolio and the success in marketing this product or portfolio. An ex-post analysis however will not be all that relevant in the hyper-dynamic mobile phone market and can of course not be conducted with regard to mobile devices that have not been introduced yet. Instead, it is necessary to apply a methodology that enables for forecasting future choice behaviour.

The methodology
When a consumer purchases a new mobile phone, he or she makes an assessment of the different models available. After reviewing the different options the model that best fits the individual's needs and wants is chosen. Mobile phones, and especially the devices introduced nowadays, are made up of many different features. The aim is to identify those features that consumers are attracted to in order for the manufacturer to incorporate the right features in its products. To estimate and realise the potential of Sony Ericsson's product line one has to anticipate the specific choices that millions of individuals all over the world will make. Simply asking the consumers regarding their future choices (stated behaviour) will not be sufficient as previous studies taught us that mobile phone users tend to be inaccurate even regarding their current behaviour.

A more promising approach is to understand the decision process and decision rules of the individual as well as to understand the drivers behind those decision rules. Accurate and reliable information regarding the sensitivities to - and perceived value of - the features of mobile devices is necessary to define an optimal product line. A thorough insight enables marketing intelligence to make a valid estimation of consumer needs, even if developments in the telecom market turn out slightly differently to what was previously expected; something that is very likely in such a dynamic market.

A trade-off approach such as conjoint analysis is a suitable methodology. Such a methodology does not focus on testing the attractiveness of products, but of product features / characteristics (including brand and form factor) relative to price. Thereby, we learn in detail the extent to which each of the characteristics impacts on choice behaviour and the extent to which each characteristic can add value to a mobile phone. Especially since this added value and the relative impact on behaviour is quantified. The study focused on the value of features of sophisticated mobile devices. Besides core features such as brand, price, form factor and size, more advanced add-on features such as type of camera or type of music player were included.

There are several types of conjoint available (e.g. Adaptive Conjoint Analysis, Conjoint Value Analysis, Choice-Based Conjoint). When deciding on which type of conjoint should be applied it is best to choose a method which adequately reflects how buyers make decisions in the actual marketplace. This includes not only the competitive context, but also the way in which products are described, displayed and considered . Of the types of conjoint analysis available Choice-Based Conjoint (CBC) analysis mimics actual purchasing behaviour the best. The problem faced by many researchers is that standard CBC set-ups do not allow the researcher to include more than 5 or 6 different product attributes in the design of the study. If more are included CBC questions can overwhelm respondents . However, a mobile phone is made up of many more relevant product attributes. The idea of resorting to Adaptive Conjoint Analysis (ACA), which is often used to deal with products that consist of many attributes, was rejected, as a typical ACA interview is a very academic exercise which does not mimic reality at all.

A commonly applied solution to overcome this shortcoming of CBC is the so-called partial profile design. With partial profile designs, each choice task consists of a randomly chosen subset of the attributes. Across all tasks and respondents a much larger list of attributes is evaluated. It is assumed that respondents can evaluate the partially profiled product concepts while holding constant all attributes not represented . In other words: the respondent is asked to make an evaluation and choice under the ceteris paribus assumption. Although this assumption is well known and often used in the scientific world many consumers are less experienced with it and have trouble in applying it.

As an alternative, it was decided to develop and use multi-step conjoint modules. The assumption was that consumers CAN handle a ceteris paribus assumption as long as each combination of features shown makes up an holistic concept by itself. Analysis of decision-making processes reveals that decisions regarding consumer durables consist of several phases. First, the consumer makes a choice regarding the brands that will be considered, the amount of money s/he is willing to spend and the kinds of form-factor that will be considered. After this initial choice process a limited number of products are left. Which of the products from this shortlist will be purchased is then determined by comparing other characteristics, such as technical parameters and add-on product features (e.g. an mp3 player).

To ‘re-enact' this process in the interview, two separate conjoint modules are designed: the first to cover the initial step in the decision-making process, the second to cover the last step. As a consequence consumers regard the concepts in each choice task as holistic and intuitively recognise the steps in the purchase process.

The specific analysis methodology used here is part of SKIM Analytical's Feature Value Modelling technique (FVM). This methodology allows the researcher to integrate the different modules on an individual level and thereby enables him / her to provide a truly holistic solution to product development issues. Incorporating information regarding preferences (established in other parts of the questionnaire) in the utility analysis makes the results more robust.

The outcomes
Analysis of the integrated choice-based trade-off exercises reveals on an individual level the ‘rules of decision' that are applied by the consumer. Not only do the preferences regarding the features become apparent, but also the relative weight each of those preferences has on the purchase decision. Hence one can determine how products can be made more attractive and how each alteration by itself or in combination will attribute to the perceived value of the products. A manufacturer can thereby make the trade-offs necessary when developing a new product. Also, brand equity can be measured, as well as price sensitivity (how effective will lowering the price of product X be).

The relative weight of each of the features is expressed in utility values. As utility is a rather abstract measurement unit the price sensitivities, which are also measured in utilities, can be used to ‘translate' the utility value in US dollar value. Therefore, it is necessary to establish the ‘exchange rate of utility and dollar' via regression analysis. The translation can be expressed by the following formula:



An example of the outcomes of the results can be seen in the chart. As the real outcomes of the study are regarded as contributing to Sony Ericsson's competitive advantage, all the results shown here are fictitious.

Fictitious example of the added value represented by each feature


Note that the chart shows the added value of each feature perceived by the aggregate market. Heterogeneity thereby has an impact on the results. Aggregate forecasting leads to an under-estimation of importance of attributes that cause heterogeneity. For example, some consumers may prefer a small mobile phone so that it is easier to carry. Other consumers prefer a larger mobile phone as it appears more robust. These two opposite opinions can of course not be combined to one "average" opinion that the size of the phone adds no value. However, when forecasting choices at the aggregate level the mathematics would lead to this incorrect conclusion .

As the utilities can be calculated on an individual level (by means of Bayesian statistics), the heterogeneity of opinions can therefore be taken into consideration. In other words, a chart like the one presented here can be created for each individual respondent. Interpreting a 1000 of these charts is of course not feasible, but the individual utilities can be loaded into a simulator .

Using the simulator not only corrects the results for heterogeneity, the product offer of competitors can be taken into account as well. Applying the simulator also makes study results dynamic. If one of the players decides to alter its strategy (e.g. lower the price of a less successful phone, or even introduce a new mobile phone altogether) the impact it will have on the share of choice for that mobile phone, the total portfolio of the competitor as well as the portfolio of Sony Ericsson can be estimated rapidly. Moreover, the simulator generates insights into substitution effects and cannibalisation effects.

Fictitious example of a price reduction for a mobile phone


How to assess the value of mobile services?

The second study focused on the value of advanced mobile services, such as MMS, video-telephony, location based services, email, etc.

Whether or not a service is valuable depends on three factors:

1. The perceived attractiveness of the service: to which extent does the service trigger the consumer and how often will the service be used?
2. The fee that can be charged for the service
3. The willingness to spend money on the service (i.e. the capitalised value of the first two factors)

The methodology
As the conjoint methodology served us well in the first study, the initial idea was to apply conjoint modules again. However, after analysing and modelling the problem in more detail, some limitations of a conjoint design became apparent.

In a standard conjoint approach preferences can be measured and choice behaviour can be simulated. The simulated choice behaviour reveals which product or service best fits the (perceived) requirements of the consumer, i.e. which product offers the most value for money. Volumes of the use of the product or service cannot be taken into account. This was not a problem when dealing with new mobile phones, as it is safe to assume that the large majority of consumers buy only one new phone at a time. This is however not the case with the services. Cost is not an independent attribute, as it depends on the volumes of use of the services (e.g. the number of MMSes sent). Furthermore, the willingness to spend extra money on a service is not only based on the perceived value and perceived benefits of the services but also on the total budget the consumer is prepared to allocate for mobile services and on the disposable income.

The choices that respondents make thereby consist of simultaneous trade-offs:

between each service and its (variable and fixed) costs AND
between different services AND

between the services combined and total costs AND
between total costs spent on mobile services and the total budget available

It became apparent that none of the types of conjoint (CBC, ACA, CVA) would be suited for this task.

In order to retrieve the information needed, a new trade-off method, named Volume Budget Trade-off (VBT) was developed. This new method completely overcomes the above-mentioned disadvantages of a standard approach and enables the 4 different trade-offs described above by showing the consequences of intended behaviour (including the volumes of use) in easy-to-grasp terms and offering the possibility of fine-tuning intended behaviour.

The mobile phone budget ‘today' and the current volumes of use of mobile services serve as a benchmark. Therefore, it is necessary to establish (on an individual level) average spending on mobile telephony, as well as the number of SMSes sent, the number of calls and the length of calls.

Previous studies on behalf of providers of mobile services taught us that stated behaviour regarding the use of services can be unreliable. The estimates of consumers regarding number and duration of daily calls tend to be very inaccurate. When consumers are asked for actual monthly spending, the results are more realistic. However, this has a severe drawback: although one knows the total monthly spending, it is not likely that one has reliable results regarding the costs of the different services (SMS, WAP etc.). Therefore, we ask the respondent during recruitment to bring a recent phone bill with him / her to the interview. The specified phone bill helps the respondent in answering the questions regarding usage and budget.

The next step is to introduce a new service by means of the stimuli. All interviewers are thoroughly briefed and provided with FAQ files to make sure the respondent grasps the concept. The fixed cost (per month) and variable cost (either per minute or per unit, whatever is appropriate) are introduced as well. After some exploratory questions aimed at establishing spontaneous interest, the respondent is asked to indicate how often he or she would use the specific service.

Once the price of the service is known and the intended volume of use is known, it becomes possible to calculate the cost for this service. Since the current monthly spending on mobile telephony is also known, the impact on monthly spending can also be calculated. The respondent is then informed about the ‘new' monthly phone bill and is offered the possibility to adapt his/her intended behaviour. The respondent is offered not only the opportunity to adapt the use of the particular service discussed at that point in the interview, but s/he can also fine-tune the use of other services (make trade-offs). By offering each service at various prices, insight can be gained regarding price sensitivity. The questions, which are administered via a computer, are looped so that the respondent can adapt until the intended usage of the different services and the consequent ‘new phone bill' are exactly tailored to the wants of the individual respondent.

The figure below shows the VBT procedure in a nutshell.



The outcomes
The results of the Volume Budget Trade-off (VBT) module provide an easy to grasp insight into intended behaviour. The method generates insights into the volume of use per week or per months for each of the surveyed services. The fact that the implications of intended behaviour on spending are shown instantly proves to be a valuable feature of the technique. Many respondents opted for adapting their intended usage after knowing how much or how little their spendings on mobile telephony would increase or decrease. For example, after being confronted with the monthly cost of video calls, more than 50% of our sample chose to adapt. This fine-tuning guaranteed much more deliberated (and consequently reliable) results than would have been the case had the results been based on stated intentions alone.

As services are introduced at different prices, price elasticities can be deduced from the analysis as well. These price sensitivities provide input for determining an optimal price.

Fictitious example of the price sensitivity regarding online gaming


In addition, per individual the willingness to spend on different services as well as the total budget allocated for mobile services can be calculated easily by multiplying volumes of use and prices. Thereby, the mobile phone budget ‘today' can be compared to the mobile phone budget ‘tomorrow'. Note that the willingness to spend on a service for a consumer is dependent on the benefits the service enables for the consumer. As totally new benefits are enabled by the new services, in this particular case benchmarking against the current mobile phone budget is rather deceiving.

Fictitious example of the willingness to spend on (new) mobile services


Conclusions

The developments in the telecom industry over the last 2 years urge both manufacturers and operators to assess the merits of new technology, and products and services derived from it, even more carefully. The importance of understanding and anticipating the preferences and behaviour of the individual consumer will increase. Especially as the stakes are much higher now!

To make a valid assessment of the value of new products and services it is necessary to observe and analyse the decision rules the consumer applies when deciding on a purchase.

Choice-Based Conjoint analysis can be a valuable tool for understanding feature preferences and the relative weight each feature has in the purchase process of mobile phones. To ensure the quality of the data it is imperative to mimic reality as much as possible during the interview process. Therefore one must offer the respondent holistic concepts when testing elaborate mobile phone concepts in the field. Standard (partial profile) conjoint designs fall short in this aspect. The Feature Value Modelling approach offers the researcher the flexibility to overcome this problem. Applying Bayesian statistics enables an analysis on an individual level and thereby effectively deals with heterogeneity in the dataset.

The Volume Budget Trade-off methodology is an excellent alternative to conjoint analysis when volumes need to be taken into account. As preferences regarding mobile services are ultimately expressed in terms of volumes of use, this new technique is the most appropriate tool here. In addition, it allows the researcher to estimate price sensitivities, substitution effects, willingness to spend and the total budget allocated for mobile services.

Epilogue

One cannot directly ask consumers to develop or design new products or services. They are not the experts - we are. However, consumer information is critical.

These studies help us to understand the drivers and barriers and realise the impact they have on acceptance and adoption of new offers on the market. The results from these studies are fully integrated in Sony Ericsson's Mobile Communications' process of product development, portfolio planning and market communication. Currently, consumer insights are treated as important as technical aspects when developing new products / product lines and services.

Sony Ericsson is sharing the survey outcomes with some of the providers of mobile services and content providers, as the adoption and success of mobile phones and other mobile devices equipped with new features depends heavily on the possibility of using these features for new mobile services. Hereby, Sony Ericsson makes its contribution for getting the industry back on course again and reaffirms its position as a leader in the industry: not only from a technological viewpoint, but also from a (consumer-driven) marketing viewpoint.

Literature

Hoogerbrugge, C.M., Huisman, D.,
Estimating the potential for a UMTS related service
ESOMAR Publication Series, volume 240, 2000

Kotler, P.,
Marketing Management
Prentice Hall, Englewood Cliffs, 1991

Orme, B.,
Which conjoint method should I use? (revised edition)
Sawtooth Solutions, Winter edition, 1998

Orme, B.,
The CBC Advanced Design Module Manual
Sawtooth Software Inc., Sequim, 1999

The authors

Annica Towliat M. Sc. (B.E), Senior Product Research Manager, Product Research Global Product Planning, BU GSM/UMTS, Sony Ericsson Mobile Communications AB, Sweden.

Jürgen R. Warnecke M. Sc. (B.E.), Senior Project Manager, Co-ordinator of the Durables & ICT division, SKIM Analytical, the Netherlands.