The Case for Massive Automation, Now
Overlogix does automation. We like it, and believe it to be one of several decisive factors in modern business. Your mileage may vary. As always, caveat emptor.
Automation is simply programming (software) with definite business purposes, mostly having machines do well-known, well-understood, boring, repetitive tasks, over and over, exactly the sorts of activities people prefer not to do. Computers can be used as robots, and will tirelessly work 24 x 7 x 365.
Computers progressively get more powerful and cheaper as time goes on. There are no issues with employment laws, no HR needed, no lawsuits, no complaints, no whining, lower risks. With all those advantages, what's the problem with adopting large-scale automation? Why isn't it already priced in to business?
As readers are no doubt aware, automation is already in wide use. We all get automated e-mail communications from organizations regularly; in fact, management of such e-mails takes up significant amounts of time. Inboxes can get completely inundated with communications from companies and organizations, so much so that the vast majority of organizational, automated e-mails contain links so consumers can unsubscribe at will. We have to have that, so the few e-mails we actually need to read don't get lost in all the junk.
"I sent you an e-mail!" used to mean something. Now, it may not, and SPAM is one of several possible factors diluting communications. We at Overlogix, for example, have been extremely diligent, for years, pruning ourselves from e-mail lists, and still have to do it several times a week. This is due to fairly promiscuous e-mailing policies on the part of a large fraction of merchants and organizations.
E-mail lists are one obvious, relatively simple form of automation. We note, without prejudice, that companies often include people in e-mail lists by default, and shift the responsibility for managing e-mail preferences to the recipients, i.e. their customers. The process has become automatic and, frankly, rather mindless.
Some online merchants offer choices in what categories of e-mails customers want to receive, with varying degrees of success. Sometimes, there are only a few categories, automatically presented at the press of a button, and the choices are fast and simple.
Other times, there are so many choices offered that the customer has to spend significant time (and can experience significant irritation) reading and making choices. Often, a login is required before the preferences can be changed. Sufficiently annoying e-mail preference changes can motivate customers to choose to opt-out entirely, destroying business value instead of supporting it.
We've done plenty of opting out. So have many readers of this article. In our case, the threshold of annoyance has become much lower over the years; it only takes one unwanted communication to hit the SPAM button and have done with the annoying robot, a bit like swatting mosquitos. There are a number of firms from which we won't accept any communications at all, due to past, transgressive, automated conduct.
The above example is so common, most of us don't think too much about it. However, business owners, especially start-ups, should take this example to heart. In the beginning, customers are difficult to find and even more difficult to retain.
Annoying them with too frequent, too aggressive, assumptive, wrong or otherwise annoying communications (read: bad automation) is a classic business mistake. Buying mass e-mail lists and relying on SPAM to grow business is even worse, since it is likely to cause entire domains to appear in SPAM lists.
The point is hopefully clear: automation needs to be done thoughtfully and judiciously, with a light touch, and more than a little empathy for those who consume the automation. This goes for automated sales, billings, payments, etc. The business case needs to be clear, the irritation factor must be considered, and a merely bean-counter approach must be avoided. To us, this is the principal risk to the use of automation: thoughtless, poorly designed robotics can destroy business instead of support it.
The other factors to consider in automating business are rather obvious: cost, culture, time-to-market, regulations and security. All of these should be classed in the category of trade-off decisions. We'll devote a little time to each one.
Cost: Automation, especially customized automation, can be expensive. Pre-made applications for most common business functions already exist: online stores, including check-out and payment solutions, have been around for a long time. Most of these are readily adaptable to a particular business, especially online retailing. Most work well and are not prohibitively expensive, since the retailer or a hired developer does the heavy lifting.
More complicated business functions, such as customer relationship management (CRM) software, can also be automated using pre-created applications that are then customized. These systems can get large and complicated, requiring more time and effort (and hence cost) to customize.
At the top of the cost category are 100% customized solutions, developed from scratch. These range from simple, rather "dumb" scripts to major applications, such as air traffic control applications. They are expensive in the sense that custom development is slow and relatively costly to produce on a per-function basis.
Culture: Laws and regulations vary over different areas and times. Culture varies even more. Mass cold-calling, for example, sometimes automated, is common in the US, frowned upon in most of Europe, and forbidden in a few places. For a business to be acceptable and to flourish, culture should not be ignored. Flaunting cultural norms may work at first, but eventually, it catches up with the perpetrators. In general, automation should harmonize with the culture in which it works, not against it.
Time-to-market: See cost above. It's always worth the effort to shop for a pre-existing solution to a particular business case, such as online stores, than to commission a custom solution. If no such solution exists, then custom solutions may be the only choice. Custom software is relatively expensive and takes a long time to produce and test, something that must be considered early on the planning of the business.
Regulations: A few industries, such as banking, insurance, medicine and pharmaceuticals, are heavily regulated almost everywhere. Other industries may be more lightly regulated, however, compliance with regulations is still important, and the software can help a great deal with compliance, both in terms of accuracy of results and timeliness in reporting.
In fact, mitigating compliance costs is one of the principal motivations for business software. Since compliance is non-negotiable, often codified into law with legally significant penalties attached to non-compliance, software can be written to enforce compliance with regulations and avoid exposure to expensive fines and legal sanctions.
Security: Customer data is precious and valuable, and data breaches, where such data is either stolen or compromised, must be avoided, almost at all costs. Automation software must be written with data security in mind from the beginning. Regulations intended to protect private data are already commonplace and increasing in both number and detail. The reputational damage from a single data breach can literally destroy a company. Software plays an increasing role in managing large lists of customers, ensuring data security and privacy, and enforcing regulatory compliance.
New business functions: Not included in the categories above would be routine automation of business functions that do not yet have shrink-wrapped, pre-created solutions. It is this category of business automation that interests us the most. Can a machine be programmed to do something, instead of hiring people?
Often, the answer is yes, particularly if the task is repetitive and mechanical. Automation then becomes a mindset, a necessity, instead of a nice-to-have. If a machine can be made to do a given business function, at scale and securely, then such programming is worth serious consideration. With time, new software solutions can be configured and packaged to be marketable products themselves, generating income instead of incurring costs.
In conclusion, the case for automation can be justified by the following factors:
We personally feel like solutions to this last concern are desperately needed in the software marketplace, and practically screams of the tremendous importance to data security of ever increasing levels of software quality. To us, software should be written so that it cannot be breached.
How much of your business can be automated? Have you shopped for pre-made solutions to your own routine tasks? Does your business management plan include automation as a first choice instead of a last choice? How do you evaluate the quality of the software you intend to use?
If no solution for your business needs exists, but you suspect it could or should, contact us about it. You'll be glad you did!