eBusiness Weekly
BusinessWeekly Last Word
When it comes to public relations, Cassava Smartech Zimbabwe and its EcoCash operation showed this week exactly what not to do, having already given a warning to other firms reliant on IT systems that pre-installation testing needs to be enhanced.
Cassava started right in its planned upgrade of its EcoCash platform. There was adequate and extensive advance publicity and advertising warning the 10 million EcoCash users that the system would be down for 24 hours from Saturday night to Sunday night.
But then the trouble started. After the migration of platform, the system simply was not working. And for the first two days Cassava just issued inane statements that the migration had been successful and the new system was being monitored, as if its customers were a bunch of dummies who could not be trusted with the truth.
Even the Wednesday statement, while at least admitting things were far from perfect, was short on detail and short on timelines for restoration of full service.
Cassava are not alone. FBC, a large modern bank and the major Zimbabwean-owned success of banking liberalisation, decided to upgrade its internet banking system over the Heroes Holiday weekend. It took a month to get it working again and customers were not exactly swamped with information on what was going wrong and when everything would be fixed.
Everyone in business knows that when it comes to IT systems, the software suppliers and our own in-house IT managers and implementation team are cheerfully optimistic that everything will work according to plan. Perhaps only the cheerful study IT at higher levels or perhaps the smiles come with those degrees.
And most of us know from bitter experience that it almost always takes longer and costs more than we were told unless we have taken some fairly stringent precautions and possibly have a competent and cynical project manager thinking through all the things that can go wrong beforehand.
One obvious problem is inadequate testing. This can be expensive in both cash and manpower. Looking at the Cassava case, there must have been some testing beforehand but clearly not enough. A full test would entail something along the lines of finding or renting enough server capacity to mirror the entire system, performing the changeover on that mirror, and then having a couple of hundred unallocated Econet lines activated for Ecocash and getting IT staff to physically see what they could see and do with their phones.
This would be expensive and would take time. But a decent sample database and the full application could have been mirrored more cheaply and still fully tested. The EcoCash changeover was not an emergency operation.
Cassava, no doubt on very good advice, decided that an upgrade was required but if it been done a couple of weeks or even a couple of months later we doubt there would have been any disaster on the old system.
FBC could have done the same sort of testing although probably used a vastly smaller sample of their data base on the mirror testing.
This is one of those management cases where the example given by the best military minds can be used with advantage. Modern weapons can be fiendishly complex and push technology envelopes to the maximum. The American procurement system is so open that we know now how they avoid some of the serious errors that hit them in the past. Suppliers promise the earth to gain a contract; eager beaver officers cheerfully make it clear that the new thing is just what they want for Christmas. But the US Government now has a special independent office that checks on progress and adherence to promises and the branches of the US defence establishment have all put in special testing units that are required to test extensively and realistically. These are normally run by officers who are not optimistic cheerful types, but hard-faced men and women who measure carefully. And finally the sign-off needed before a production run is authorised is done by a perfectionist. We know less about how say the Russian air defence command or the Chinese Peoples Liberation Army does this, but it cannot be much different even if less public.
Even in operations, the military mind tends to go for extensive training and rehearsal under realistic conditions. Taking a Second World War example, the D-Day invasion of France by the Americans, British and Canadians. Troops were trained and retrained and forced to rehearse realistically. Several hundred in fact died in accidents in these rehearsals. But on the day a lot went right. A lot went wrong as well, but because commanders and junior officers and even the troops had been so heavily trained it was possible to make changes to the plans.
Incomplete bombing saw German artillery batteries still functioning, so admirals ordered their cruisers dangerously closer inshore to engage in accurate gun duels; bad weather saw task force commanders beaching tank landing craft instead of launching amphibious tanks at sea; even second lieutenants in charge of a troop of tanks switched targets and switched roles for specialised tanks. So the invasion was a success.
The second error by Cassava was not explaining to their customers from the start what had gone wrong and why, what was being done to fix the system and giving an indication of timelines. The lack of sensible information, or even an awareness by Cassava managers that something was seriously wrong, probably irritated customers more than the actual problems in the first place. People want to be treated as rational adults and a wise company does this.
EcoCash is in a unique position, handling 85 percent of all financial transactions in Zimbabwe and 99 percent plus of all mobile money transactions.
It achieved this enviable near monopoly not by some decree but through free-market competition by offering a service that had exceptional levels of reliability at very low transaction costs and near universal acceptance. Even those who use other networks for communication almost always have an Econet line as well for EcoCash. While EcoCash has 10 million users in a population of 14 million, it is not used by babies or small children, and so its user base includes almost all adults and many teenagers.
That sort of position requires responsibility. Already there are many customers now wondering if they have put too many eggs in one basket and are looking at similar systems offered by Net-One and Telecell. Cassava must start taking into account pressure to switch mobile money into something more like the ZimSwitch model, where all the banks have agreed to use compatible technologies, the same basic swipe machines and a common interface linking all their databases.
Even if EcoCash is working perfectly today, Cassava’s reputation has taken a knock. But we all learn from its mistakes. Test and test, rehearse and rehearse, and then trust your customers and be open and transparent.
