Lessons learnt: a more nuanced market segmentation between on grid and off grid

Whereas M-KOPA addresses the off-grid customer base with solar home systems, Kenya Power and Lighting Company (KPLC) addresses the on-grid customer base, providing power through traditional grid infrastructure directly into the homes of those who have been connected. Much of the discussion around energy access is focused on these two categories, dividing customers and households purely by grid connections. One of the central findings of M-KOPA’s Shell Foundation funded market analyses is of a much more nuanced market segmentation between on grid and off grid; M-KOPA identified three additional and distinct groups of customers that do not fit cleanly into either category. These three segments have been titled “under grid”, “idle grid”, and “bad grid”, and each has slightly different customer pain points that M-KOPA and companies like it can address.

The needs/wants of on grid and off grid customers are broadly understood already, and research did not uncover significant new information about these two segments. However, research into the other three presented some new insights, and significant market opportunities which are set out here.

Table 1: five customer segments


Figure 1: Percent of Kenyan households in each grid connection category


The under grid segment

The Energy Institute (EI) at the Haas School of Business (UC Berkley) undertook research in 2015 on rural electrification efforts in Kenya and coined the term “under grid”. The term refers to households within connection distance of a low voltage transformer (600 meters in Kenya) but not connected to the grid. The EI research found that only 5% of households and 20% of businesses within 600 meters (in communities where a low voltage transformer been installed within 5 years) had grid connections via the transformer. There is a significant gap between the number of Kenyans who in theory could connect to the grid, and the number that has.

Subsequently, schemes to reduce connection fees, or amortise fees over the first few years of electricity bills have increased the percentage of connections in these communities substantially. More than 60% of Kenyan households are now connected to the grid. However, the problem of the under grid is persistent – according to recent M-KOPA research, 11% of the population can still be classified as under grid, and this is likely to increase to 16% with the newest round of grid extension efforts, as planned grid extension efforts bring new transformers within striking distance of previously fully “off grid” customers.

As illustrated in Figure 2, the under grid population in Kenya is disproportionately lower income. Amongst households above the poverty line, more than 80% are in the “floating class” (making between $2 and $4 a day). These households are more likely to gain income from irregular sources, or from running businesses, and 50% have no lighting sources (including solar home systems or solar lanterns). This customer group shares many characteristics with purely off grid customers, and therefore, are prime candidates for solar home systems financed on the pay as you go model.

Figure 2: Distribution of floating class and middle-class households by each grid connection category


The idle grid segment

In early 2017, the Standard newspaper published an article asserting that nearly a third of the 3.5 million households KPLC claimed to have connected to the grid in the most recent Last Mile Connectivity Programme were inactive. This meant that they hadn’t topped up their pre-paid meters at any point in the first year after connecting. Their meters were preloaded with 30kWh

of credit, meaning that 1 million households were using their KPLC connection for a single lightbulb or less during the period since connection. This group of customers typifies the idle grid segment.

The idle grid segment shares many features of the under grid segment, except that they have grid power already connected to their home. Generating electricity demand is the key problem for energy suppliers seeking to serve this segment; idle grid customers rarely have access to affordable appliances and they lack sufficient disposable income to afford large electricity bills (1).

These two features make it difficult for idle grid customers to gain the full benefit of their electricity connections, or to generate revenues for connection providers.


The bad grid segment

One third of grid connected households have connections that work less than half the time (2) representing more than a million households in Kenya without reliable power. Demographically the bad grid customer segment is between the good grid and idle grid segments. Although they have higher incomes and are more likely to earn a regular paycheck than idle or under grid segment households, they are disproportionately peri-urban and more likely to cite unstable power as a reason they don’t own more appliances.


Visualising the segments

As illustrated in Figure 3, good grid is most pervasive in urban and peri-urban households (largely a result of the quality improvement efforts undertaken around Nairobi in recent years). Bad grid and under grid customers are found in peri-urban areas, while the idle grid segments have a substantial urban presence (largely indicative of the low incomes of customers in slum areas impacting their ability to afford appliances and electricity).

Under grid and idle grid customers have similar income levels, while bad grid and good grid customers enjoy higher incomes and, therefore, are more lucrative to service than either the idle or the under grid segments.

Figure 3: Percent of households urbanised by each grid connection category


Figure 4: income sources by grid connection category


Regular incomes are common within the good grid segment since urban centres including Nairobi, with their larger employment markets, make up a substantial percent of good grid households. Irregular incomes are common within the other segments, with self-employment (through small business or agriculture) making up the largest share in the under grid segment, followed closely by the idle grid segment.


Conclusions on market segments

Each category of grid connectivity suffers from different challenges that companies have the potential to address with well targeted product offerings.

  • The bad grid segment primarily needs more reliable sources of energy. This could come from solar home systems, running in parallel with the grid (indeed 17% of M-KOPA customers already have a grid connection in their homes), or from grid backup systems that can provide power during grid power outages.
  • Idle grid segment customers are unable to use their grid connections beyond lighting. They need highly efficient appliances available either at a subsidised price, or through accessible financing schemes, to enable the affordable use of energy (including for productive purposes).
  • Under grid customers can be targeted to be brought onto the energy ladder, starting with a solar home system and eventually graduating to a full grid connection. They also would benefit from affordable financing for grid connections and the wiring and appliances needed to effectively use electricity in the household.

Each segment is unique and presents different commercial opportunities for the private sector. Each also requires a different and nuanced approach from government, NGOs, and donor agencies to help them realise the benefits of energy services.


(1) 40% of idle grid respondents in an M-KOPA survey said they need to limit their consumption because of the cost of electricity.

(2) Afrobarometer study, 2017.


This text is an excerpt of M-Kopa Labs’ new “Lessons Learnt report”. M-Kopa Labs was established in 2016 to accelerate the company’s ability to produce innovative, larger capacity appliances capable of interfacing with off-grid energy. In developing new devices, M-KOPA has also been able to explore and uncover additional understanding of its markets and the energy-use of existing and future off-grid customers.


Download the full “Lessons Learnt report” here.