Expanding Markets: Mathew Miller of OSIsoft
There are four basic growth strategies for firms to increase their revenues. These are penetration, geographic, horizontal, and vertical. Each growth strategy has its own internal characteristic growth rate, beyond which growth is achieved in non-economically efficient manners. Selecting the appropriate growth strategy and growth pressure for a firm requires determining competitive strengths and managing the value proposition.
This case study examines how OSIsoft has utilized these growth strategies after providing basic definitions of the four generic growth strategies.
A penetration growth strategy increases the attack within a chosen market to increase the market share. A geographic growth strategy is as it sounds: increase revenue through increasing the territory. Often, as penetration strategies begin to peter-out, firms resort to geographic growth strategies in expanding to new cities or nations. Penetration and geographic strategies both leverage readily identifiable core strengths of the firm; that is: the firm understands the product value proposition, competitive landscape, and customer demand drivers. Also, the firm can leverage its brand identity, past success stories, and current operational strengths. Early within a company’s or business line’s lifecycle, managers will select penetration and geographic growth strategies as the best means leverage their traction and improve revenues. At some point in the product’s or firm’s growth, other levers become more efficient in driving revenue growth.
More daunting challenges face a firm as it embarks upon horizontal or vertical growth. Horizontal growth refers to selecting a broader number of vertical markets to attack. Horizontal growth requires an increased marketing effort because customers within the new vertical must be introduced to the product and its firm. Vertical growth refers to selecting other positions within a particular vertical wherein the company can create and capture value.
OSIsoft has experienced each of these growth strategies including penetration, geographic, and horizontal. OSIsoft has sold over 6000 systems in over 80 countries to a multitude of industries such as power plants, chemical facilities, pulp and paper mills, food and drug manufacturing plants, mining and materials facilities, and network monitoring facilities. As of the end of 2001, Dr. Kennedy, Founder and President, has led OSIsoft to $63 M revenue after 22 years of operation.
In the past 12 months, I found OSIsoft at two very different conferences. At Distributech, OSIsoft highlighted its PI System product to the utilities vertical (www.osisoft.com). At the Embedded Systems Conference, OSIsoft highlighted its ECHO product to the hardware market (www.osisoftecho.com). When OSIsoft presented ECHO at the Embedded Systems Conference, they were executing the fourth growth strategy and increasing their markets vertically. Why did they select this growth strategy now?
OSIsoft’s fundamental value proposition resides in capturing and managing time-series data. Unlike business applications that have adopted n-tier architecture with a relational database foundation, time-series data coming off a machine is more efficiently and effectively stored in a temporal database. Temporal databases differ fundamentally from relational databases in reflection of their purpose – to store a long data-stream for later retrieval. For example, business systems must capture state variables such as those concerning a financial transaction or the length of a phone call. In monitoring machines, data coming from the machine might be arriving every second or millisecond. Like a flight recorder, the temporal database must capture all of this data. Unlike a relational database, not every data point in the data stream must be captured. Using algorithms such as swing-door compression, a temporal database can look at values and determine if the data is changing or remaining linearly constant. If the data is linear, only the first and last data point needs to be stored while the intermediate values are extrapolated upon demand. The effects of using temporal database over a relational database can be significant in both performance and storage. For example, monitoring a machine and capturing 10,000 data points per second would use 8.5 GB of data storage in one year with a relational database, but a temporal database can usually hold this same data content with only 0.8 GB.
To attack their original market, OSIsoft sells their temporal database as a component of the PI System product. PI is a business application in the “real-time information management” product category, recording and managing
real-time data coming from multiple sources in the form of data streams. PI has as a foundation component a “digital historian” to manage the temporal data. For the Utilities vertical, PI represents a means to capture all the
data coming off a generation turbine or transformer in real time, thus explaining their presence at Distributech. ECHO is composed of a digital historian too, but designed specifically to be embedded within a particular
machine in its own database. ECHO is positioned as an enabling technology for OEM manufacturers to improve their value proposition to business customers. This represents a Vertical Growth Strategy because now OSIsoft’s
products are at both the business application level and at the origin of the data-stream embedded within the machine itself.
A number of industry trends indicate that machines are becoming more intelligent. Microsoft and Sun have made significant investments in their Windows CE and JINI for the embedded systems market, respectively. On the
hardware side, Motorola, TI, and others have developed a number of chips and solutions to monitor appliances, heavy equipment, and remote assets. With a digital historian product line, it makes sense that OSIsoft would
recognize a potential industry trend to embed a temporal database within the equipment for collecting and managing time series data.
When OSIsoft determined to leverage their core strength in temporal databases to the embedded systems market, they had to make some changes according to Mathew Miller in OEM Sales & Marketing for OSIsoft. For instance, the embedded temporal database couldn’t be sold under the PI brand. PI has a very large value proposition as an enterprise business application. Selling an embedded system with the PI name would confuse or lower the perceived value of PI as a business system. OSIsoft solved this dilemma with the ECHO branding. Also, the target customer changed. Rather than selling OSIsoft’s products to the end-customer business, OSIsoft sells ECHO to equipment makers, appliance manufacturers, and other upstream manufacturers. With this new target, these manufacturers manage the end-customer value proposition, not OSIsoft. Likewise, OSIsoft’s selling story has changed. Rather than focusing on a number of case-stories explaining how a customer has increased profits using a temporal database, OSIsoft now focuses on the ability of a manufacturer to sell more units or command a higher price.
When asked what the largest barrier was to purchasing OSIsoft’s product, Mr. Miller’s response was that many customers don’t understand the value proposition of temporal databases. After the industry has preached the virtues of relational databases for over two decades, it’s hard to counter biases with alternative truths and the value of temporal databases. This difficulty is more of a symptom of general consumer biases than anything specific to vertical expansion, however it is useful to note that the story that OSIsoft has learned to tell at one level in the vertical can be retold at another level.
OSIsoft is having some success in the embedded systems market and with their newest vertical growth strategy. While executing this strategy, OSIsoft has not forgone its current markets or past areas of growth, rather it is adding to its market depth. As a case study, OSIsoft nicely demonstrates how each of the four generic market growth strategies can be utilized, and some barriers and opportunities associated with ertical expansion.
The May Report, TECH BUSINESS BRIEFS, July 25, 2002