Mobile network operators (MNO) have recently shown interest in offering insurance to their client base[1], described as ‘m-insurance’, offering a significant opportunity massively to scale up access to life and non-life insurance.
The objectives of each MNO may differ (it could be about driving financial returns, enhancing and differentiating their brand, or increasing ARPU) and there are a number of models in existence (such as an embedded loyalty programme, or payment by airtime deductions or mobile money, and hybrids[2]). As exciting as this opportunity is to increase the numbers of the low income (and middle and upper income) people with life insurance, there remains a danger that the promise will be unfulfilled should clients not understand or benefit from the cover or being suddenly left without cover as we will explore.
Why is m-insurance exceptional?
M-insurance, insurance sold through and with mobile operators, is exceptional compared with the more traditional insurance distribution for the following reasons:
- The potential for distribution is second to none. Kenya offers a stark example: there are approximately 3,000 insurance agents in the entire country, and yet M-PESA, the well-known mobile money initiative of Safaricom, has more than 60,000 agents.
- MNO brands are far more trusted than insurers. In South Africa, insurers were less trusted as an ideal financial service provider than informal unregistered cash loan shops (FinScope). In Ghana, 70% would prefer to buy insurance from an MNO than an insurer (MTN-Hollard demand side research).
Distribution and trust are important reasons for insurers to seek out relationships with MNOs, however, insurers (and regulators) must recognise certain challenges in these commercial arrangements:
- Power imbalance. The MNOs dwarf many insurers in terms of capital and assets and often have powerful political connections, leaving the insurer subject to significant pressure to conform to their views. As the weaker party, it is difficult to hold the other party accountable, particularly in terms of selling practices or commercial terms. In one case, even the regulator felt unable to hold the MNO accountable as a licensed agent as they did not have an internal legal team or the resources to challenge the MNO about poor behaviour.
- Multiple regulatory issues. MNOs are regulated by communication authorities, and those offering mobile money services are sometimes additionally regulated by (or subject to) Central Bank rules. When adding insurance, the additional market conduct rules from the insurance regulator adds complexity and cost, and increases the chance for regulatory arbitrage. Furthermore, the insurance regulator tends to be under resourced in many developing countries, with limited ability to impose meaningful penalties and lack, or have limited access to, legal advice. In one example, an insurance regulator felt that the MNO was too large to take on meaningfully with their cohort of lawyers, so it was able to escape any penalties. This means that without joint supervision or oversight from the communications and or banking regulator, it will be extremely challenging to enforce the regulations on MNOs.
- Client ownership and understanding. MNOs’ understanding of their clients means that they may well be in a better position to disrupt, used in the sense of disruptive innovation, the insurance market than the insurers themselves. This has unforeseen effects on traditional insurance providers and markets. This could be through better understanding of risk thus allowing superior pricing, more effective tailoring of products to need, all backed by their massive scale and brand power. Ultimately, their brand also means that they will ‘own’ the client even though, legally, the relationship should be ‘owned’ by an insurer. In South Africa, Vodacom has even bought a life and non-life insurance licence to enable their m-insurance plans without a separate insurer being involved.
Particular focus should be paid to m-insurance because of the potential to rapidly scale a product which can protect low-income households from risk, but which could also go extremely wrong without attention.
What are the challenges around m-insurance?
M-insurance creates the following particular challenges:
- Development of meaningful products that meet customer needs. Although there is the potential to effectively leverage the MNOs’ huge databases and comprehensive segmentation models to target and understand clients better and more efficiently, this is rare because it requires investment and effort. We are therefore still far away from this in practice;
- Tailoring distribution models to effectively target early adopters such as the youth who are more likely to accept remote sign up models which will drive take up. Whilst the youth may not typically be the main target of life insurers, strong cultural drivers, such as around funeral insurance in Southern Africa and other countries, often pushes take up into the youth market and the importance of the cell phone to youths could drive cell phone handset insurance;
- Creating a demand for insurance. An informed and engaged customer base that will effectively utilise the insurance product is necessary, but difficult in markets in which very few people (rich or poor) have (positive) experiences with any type of insurance. An unengaged customer base will not pay premiums, or accurately submit claims which will stop the “word of mouth” marketing potential.
- Selling insurance in a new way. Develop appropriate agent incentives and training to aid the distribution of the products – insurance is a lot harder to sell than airtime or mobile money as it is typically new to the market, often has a bad reputation and is more complex. Offering insurance “free” in an airtime loyalty programme is very different than selling insurance to existing clients through the existing products. This is particularly true in markets in which mobile money itself still has some challenges around ensuring an active base
It is clear that these challenges must and can be addressed over time and that m-insurance models offer considerable promise to reach massive scale, drive benefits for MNOs and clients and address the growing calls for an inclusive financial market.
Regulatory reflections
At the same time, there is considerable risk of regulatory whiplash should these models be designed in such a way that they do not offer client value over the long-term, breach the rules or fail. Two m-insurance examples illustrate the risk of, and indeed may frame the case for, regulatory intervention:
(i) Getting the commercial agreements right: Econet Zimbabwe cancelled its embedded insurance due to a dispute with Trustco, a technology provider, and, overnight, 1,6m people lost their life insurance[3]. The dent to the reputation and trust of insurance in the market (by clients and regulators) is serious and at systemic levels – over 10% of the total population and roughly 20% of the adult population were affected.
(ii) Client value matters. The Namibian regulator raised serious concerns about the value of an m-insurance loyalty programme due to the number of exclusions which led to extremely low loss ratios and thus undermine the flow of positive ‘word of mouth’ stories by clients[4]; claims are essentially an insurer’s shop window and overly low claims mean low appreciation of insurance.
Both these examples demonstrate how easily trust can be undermined, stall market development, and even destroy it. MNOs and insurers need to carefully consider their approach in order to enable these powerful models, but also to limit the risk of fall out.
Could a living will and code of conduct manage risks in the market?
In order to mitigate these risks, there are a couple of models that could be considered. One example that one could borrow from the banking sector may be to require a form of ‘living will’[5] (an approach that arose after the financial crisis) as part of licensing m-insurance models, where the MNOs either voluntarily agree to ensure alternative cover is available, paid or not, should they “switch” off the cover or have it imposed as a condition by the regulator (see the box). This would aim to limit the potential damage to the market should insurance be less appealing to the mobile operator who either pay the full premium or subsidises the distribution. This approach also strengthens the insurance regulator’s hand as they would have to approve these ‘living wills’ up front, rather than deal with the ex-post consequences which are more challenging due to the power imbalance between the typical under-funded regulator and the well-resourced MNO.
An alternative model that the MNOs and the GSMA, their trade association, should consider is some form of code of conduct to manage their reputation and to grow the market. This should consider issues such as driving disclosure to ensure the clients are aware of the benefits, and guidance around exclusions (in one model, for example, in a monthly term life model, a claim is only valid should a client die within that month, which appears overly onerous) and other reputation enhancing or destroying approaches.
Whilst we are excited about the potential of these models to do good through massively increasing access to insurance for the under-served and un-served, we believe that the need to “do no harm” will require a delicate balancing act between enabling these models but requiring appropriate safeguards.
Jeremy Leach is Director and Head: Insurance at BFA (Bankable Frontier Associates), a specialist consulting firm focused on innovation in financial services for the low-income market. Valuable input and comments were provided by Evelyn Stark Senior Program Officer in the Financial Services for the Poor team at the Bill & Melinda Gates Foundation.
[1] For background reading on m-insurance, see http://www.cover.co.za/short-term-insurance/m-insurance-the-next-wave-of-mobile-financial-services
[3] See http://bulawayo24.com/index-id-business-sc-companies-byo-12369-article-Econet+divorces+with+Ecolife+as+agreement+expire.html
[4] A personal communication between the author and one of the senior regulators, who did not want to be named.
[5] See http://online.wsj.com/article/SB10001424052702304211804577505011956447968.html