The title of this post seems obvious, but the topic of Last-Mile product delivery is make or break for social enterprises looking to create an impact in emerging markets. Finding a way to get their solutions in the hands of customers will cost time, effort, and lots of money. With this in mind, it is definitely worth rethinking distribution networks from right to left, even making alterations to the business model if necessary, so that SocEnts can practically set their sights on achieving scalable impact. This strategy check indicates that each last-mile segment in-place withing the network is more valuable than the return on churning basic new sales, suggesting that supplying multiple products to keep existing segments continually activated is favorable to selling a single product which may saturate certain segments. (Disclosure: this discussion addresses a simulation based on general observations of the market for sales India or on field results reported by others in the energy access community.)

Let's look at the model for a distribution network shown above and assume that the villages are roughly the same size and that their distance from the Local Hub is representative as shown. The traditional mode of dealing with such a model for delivering a product to Rural India involves many layers of middlemen, each with their own markup. This inefficiency acts counter to  delivering the desired impact for the target consumers, who are essentially priced out of the market and cutoff from supplier service. As a result, social enterprises must choose from the following two options for introducing products into emerging markets, namely, A) developing their own distribution network from scratch, or B) "piggybacking" on the distribution networks of existing brands within a target market.  These approaches are highlighted in research from the IFC as well as from Benjamin Neuwirth at  Kellogg School of Management. Solar products have been distributed via both methods, with suppliers either delivering the product to Village Level Entrepreneurs (VLE's), who are then responsible for sales within the village, or by "riding" on a "carrier" brand's distribution from the regional level. 

Whichever strategy is selected, the social enterprise must vigorously maintain and nurture the network it creates, recognizing that downstream segments (i.e. closer to the last mile) are  often more difficult to replace than those further upstream. Regardless of where offloading of the distribution responsibility occurs, there is considerable cost to developing each segment of the network, as even the costs avoided by outsourcing are purchased by some measure of revenue-sharing. Similarly, training VLE's will also incur costs along with time and effort. If we look at the case of selling a single product, assumed to be a solar lantern, we may examine possible responses to scenarios where the distribution model is stressed by stagnation:

As villages A & B start to see stagnation, the challenges of maintaining service to them become costly to maintain in light of dwindling revenue prospects. The natural reaction would be to look to the untapped "green fields" of Villages D-F, assuming that the same pattern will repeat itself, but that the new sales will justify the expansion. This logic is deceptive, because Villages D-F are further afield ("Last-Mile+") and therefore potentially more costly to service. Also, the initial assumption would be:

• Villages A-C produced an average return per village of Ravg from a total earnings of 3*Ravg

• Adding Village D might produce another Ravg of earnings, for a total of 4*Ravg 

The problem with this assumption becomes clear when considering the costs of expansion. Village D is not only more expensive to serve, but also must take up the lion's share of earnings production shed Villages A & B, all while A & B continue to incur costs for after-sales service. Even if a VLE from Village C pushes products through to VIllage E, the mathematical requirement for that sales volume to sustain overhead costs for all segments downstream of the local hub is most likely pretty steep.

Another option would be to open a new hub, but this should not be attempted simply as a means of replacing earnings from another end-of-cycle hub. The new hub will require further upfront costs to get started, which will need to be made up for in new sales, on top of the residual costs to keep the first hub on life support. 

The above scenario clearly establishes that the path to successfully reaching scale requires sales which significantly outpace cost to deliver, which, in the case of Last-Mile markets, comes primarily from setting up downstream segments of the distribution network. This effectively equates scalability with a ROI for setting up each downsteam segment, so that the ideal scenario occurs when downstream segments never close once they are set up. This is only possible if stagnation is avoided by continuing to provide new products and services across existing segments. Consider the revised alternative model below:

Let's assume that a social enterprise chooses to offer a range of products which are mutually complementary or compatible.  In this case, the enterprise can pick up or drop suppliers as needed without reshuffling the downstream network. The sunk cost of setting up service to each village is amortized over a greater number of transactions. Even in the case of slow adopters, which are one potential cause of stagnation in any village, the continued presence of service along with the variety of product offerings may result in their inclusion as customers (and impact beneficiaries!). This model may also potentially empower VLE's to increase their revenues proportionally via sales to existing customers rather than ranging father from their home village to track down new customers. Feedback from satisfied customers can directly inform new product offerings, enabling the social enterprise to engage new suppliers with confidence. Targeting such an outcome will result in a healthier distribution network end-to-end, with greater stability, resiliance, and long-term returns. With the advantage of such a robust platform, scaling through replication should be much closer to fruition.