Sunday, December 30, 2012

Customer value of smart metering: Explorative evidence from a choice-based conjoint study in Switzerland

Abstract: Implementing smart metering is an important field for energy policy to successfully meet energy efficiency targets. From an integrated social acceptance and customer-perceived value theory perspective we model the importance of customer value of smart metering in this regard. We further shape the model on a choice-based conjoint experiment with Swiss private electricity customers. The study finds that overall customers perceive a positive value from smart metering and are willing to pay for it. Further, based on a cluster analysis of customers’ value perceptions, we identify four customer segments, each with a distinct value perception profile for smart metering. We find that energy policy and management should integrate a solid understanding of customer value for smart metering in their initiatives and consider different smart metering market segments within their measures.
Highlights
► We model the importance of customer value of smart metering.
► We shape the model on a choice-based conjoint experiment.
► Overall customers perceive a positive value from smart metering.
► Customers are willing to pay for smart metering.
► There are four distinct customer segments with different value perceptions.
...
Full-size image (27 K)
Fig. 1. Customer value of smart metering.
Within the experiment customers would be willing to pay a premium of up to 9 CHF Swiss Francs current exchange rate 1CHF=1.09 US Dollars) to get their most desired smart metering product (3.04+1.55+1.30+2.38+0.93).  However, customers would also demand a discount of the same value if their smart metering product did not fit their needs. The results of our willingness-to-pay calculation for the different tariffs indicate that consumers are willing to pay more (3.04 CHF for a tariff of 11/17 Rp./kW h) to get the tariff model with the lowest risk or they would expect a discount of 3.56 CHF when forced to accept the highest offered tariff model. A possible interpretation is that they are willing to pay to avoid the risk related to high tariffs (Chapman et al., 2001; Faruqui and Mauldin, 2002; Herter, 2007; Faruqui et al., 2010). The related chance of falling into the lowest tariff seems either to be unsuitable to balance this risk or seems not to have been realized by respondents.

To compare the costs of smart metering with the willingness-to-pay we could calculate how long it takes for the costs to be amortized. In the recent published impact assessment of smart metering in Switzerland (SFOE, 2012) five scenarios for smart meter implementation and related costs were reported. The scenarios range from “status quo”, which does not foresee the implementation of smart meters and which uses the existing infrastructure, to the scenario “nationwide implementation +”, which consists of an implementation of smart meters at 97% of the metering points until 2035, a smart meter enabling infrastructure, dynamic tariffs, data collection in a 15 min. interval as well as load management for various appliances. The scenario “nationwide implementation +” allows for all of our services to be offered. Whereas the total accumulated costs between 2015 and 2035 of the scenario “status quo” would amount to 4319 million CHF, those of the scenario “nationwide implementation +” would amount to 5236 million CHF (SFOE, 2012). The costs include investment,
operating, communication costs and costs for business processes (SFOE, 2012)..... If we subtract the total accumulated costs of the scenario “status quo” from those of the scenario “nationwide implementation +” we arrive at the added costs due to the implementation of smart metering, which amount to 917 million CHF.
Today 4.9 million meters are installed in Switzerland (SFOE, 2012). If each of the households and companies who are equipped with these meters would pay 9 CHF per month for their ideal smart metering offering, this would amount to a revenue of 529.2 million CHF per year. Bringing this into relation with the additional costs for smart metering of the “nationwide implementation +” scenario, 917 million CHF, this would mean that these costs could, for example, be amortized in approximately 2-3 years.
...
The first cluster (39% of the customers) can be named “risk-averse”. The strong preference for the lowest tariff model with low risk is the dominating characteristic. Apparently, customers of the first cluster do not see the benefit of potentially lower electricity bills due to dynamic pricing programs such as CPP (Faruqui et al., 2010). One explanation could be that they have difficulties in estimating their ability to react to the prices through energy efficiency measures and thus are more worried about
the risk of an increase in their bills rather than focusing on the potential savings .... Thus, this group might not be the ideal target group for dynamic pricing programs unless these provide a risk hedging possibility.... Also customers within this segment do not seem to be willing to pay5 a higher base fee reflected by their stronger preference for the lowest base fee compared to other clusters and the total sample. Furthermore, it appears that services related to programming and steering and home security do not provide as much value to customers within this segment as to customers of other segments. As for energy consumption feedback and remote meter reading this segment’s value perception is approximately as high as the aggregated value perception of the total sample. Overall, this segment might be attracted by basic, low risk smart metering services, including energy consumption visualization and remote meter reading. The price for those additional services needs to be clearly attributed to the added services and separated from base fee. 

The second cluster (29% of the customers) can be named “technology minded”. It is dominated by high preferences for steering and programming services. Additionally, customers within this segment perceive a higher value from visualization via mobile devices (0.9) and remote meter reading with accurate monthly billing (1.44) compared to the total sample and other segments. Furthermore, the perceived value of the lowest tariff model is below that of the total sample and customers within this
segment do not seem to perceive such a high risk with a higher tariff model compared to cluster one and four. This indicates that customers within this segment might be more inclined to participate in dynamic pricing programs and demand response. Finally, customers of this segment again prefer the lowest base fee. Thus, the ideal offering of smart metering for this segment should consist of a standard base fee and additional services such as programming and steering with a high customer-perceived benefit should be priced separately to capture the higher willingness-to-pay for such services within this segment.

The third cluster (20% of the customers) can be named “price sensitive”. The most visible characteristic is the high value customers assign to the tariff model providing the lowest tariff (6/50 Rp./kW h). Although this tariff is also associated with a comparably high risk, customers within this segment seem to value the option to save on the low price tariff. Thus, this segment seems to be the ideal target group for dynamic pricing programs such as CPP. Going along with the preference for the highest tariff model is that the detailed monthly invoice is perceived as valuable by this segment. Furthermore, customers within this segment do not seem to attribute such a high value to a low base fee as the previous clusters, indicating that the base fee for this cluster is less important and could thus be augmented. Overall, customers within this segment seem to be ready to actively manage and control their energy behavior to leverage low-price tariffs. Thus, a smart meter offering enabling them to do so, e.g., dynamic pricing and services to enable demand response, might be ideal.

The fourth cluster (17% of the customers) can be named “safety-oriented”. An important characteristic is the strong preference for home security services compared to other clusters and the total sample. Additionally, customers within this segment attribute the highest value to energy consumption visualization via in-home display. Interestingly, at the same time, customers within this segment do not attribute a higher value to high tariff models. This might be an indicator that being able to react to dynamic prices with the help of energy consumption visualization might not be the only value of visualization tools. For example, it might be possible that energy consumption visualization provides value to customers by ensuring security at home (e.g., detecting malfunctioning devises with the help of visualization tools). Of further notice is that this is the only segment, which attributes a higher value to a higher base fee. Additionally, customers of this segment do not seem to be that much interested in programming and steering services and attribute a fairly small value to remote meter reading with accurate billing. Thus, an ideal smart metering offering for this segment would focus on value associated with home security services, including in-home displays with detailed consumption feedback for single devices. If remote meter reading needs to be included in the offering for allowing energy consumption feedback, it should not be priced due to the relatively small customer-perceived value. Instead the costs could be amortized through the possibility of a higher base fee in this segment.
...
Thus, a standard smart metering offering for the whole market might lead to suboptimal results as it fails to address the heterogeneity in value perceptions and thus potential differences in willingness-to-pay. Based on a value-based segmentation done by a cluster analysis of customer preferences we conclude that there are four different service bundles, which correspond in an optimal manner to the individual customer segments’ value perceptions.
Full-size image (64 K)
Fig. 2. Sensitivity analysis based on share of preferences.
Full-size image (49 K)
Fig. 3. Willingness-to-pay.
a Repower AG, Talstrasse 10, 7250 Klosters, Switzerland 
b Good Energies Chair for Management of Renewable Energies, Institute for Economy and the Environment (IWÖ-HSG), University of St.Gallen, CH-9000 St. Gallen, Switzerland
Tel.: +41 71 224 25 86; fax: +41 71 224 27 22
Energy Policy via Elsevier Science Direct www.ScienceDirect.com
Volume 53, February 2013, Pages 229–239
Keywords: Smart metering; Customer value; Cluster analysis

No comments:

Post a Comment