What is a Metric?
A metric is a value that can be measured on a scale, whether as a percentile (0% to 100%) or as an integer (ages from 0 to 120 or income from 0 to infinity). Metrics help place data into context by providing a baseline (or an acceptable average) to compare to.
For example: approximately 50% of Toronto residents live at or below the LICO (Low Income Cut-off); if an area surrounding a BIA has a greater number, a metric can show how an area either reflects a better or worse standing in comparison to another area or the city as a whole.
Metrics can also have attributes, such as thresholds (either upper, lower or both), sweet-spots (targets or ideal conditions), trends (the velocity of the variable over a span of time) and descriptive ranges that can make a number’s inherent meaning more familiar to the audience.
Real Life Examples
A simplistic example of a metric is the measure of temperature via Celsius. It has values (positive and negative), baselines and thresholds (freezing and boiling), trends (rising and falling), sweet-spots (room-temperature, or whatever to liking) and ranges (cold, mild, hot). But, as a metric, it is fairly static, aside from pushing down absolute zero and finding new states, such as plasma.
A better example is the TSX S&P500 Index as a metric. It is a select list of stocks that combine their collective value and indicates, on a whole, how the entire market is fairing. Standard and Poors – the S&P – takes a sample of select blue-chip stocks in the market, for example Research In Motion (RIM), the maker of the venerable Blackberry or Telus(T), the deregulated offshoot of BCTel, and compares their performance to create a value, or metric, that can be used by the financial industry and investors to garner a overall understanding of market sentiment. When the newsreports announce that the TSX dropped 100 points, this reflects the overall performance of only specific stocks, not all stocks. In any case, some stocks may have risen but others may have fallen, regardless if they are used within the metric or not.
Example: As the price of oil increases, the index reacts by falling as market variables dependant on disposable income, holders of household debt and airlines will fall because the price of fuel will eat away not only at their ability to make profit but their customers’ ability to purchase their goods and services or make ends meet. But, oil producers and energy companies, such as Suncor, stocks will increase, along with related industries: the manufactures of pipelines, big-rig machinery such as Caterpillar and engineering firms that design and build oil fields such as SNC-Lavalin.
It also consists of thresholds, such as the bear and bull markets where, over a span of time, the metric has lost or gained a specified percentile of value. The metric is designed to weigh a comparison of the overall market and includes a cross-section of industries (rolling up sub-metrics) to create a value that we can use to quickly create a common reflection. Often, this metric changes as companies are added and removed. During the 1990’s and 2000’s, online companies were heavily weighed in the S&P500 in comparison to other industries where before in the 1980’s computer manufactures and software companies comprised most of the value in the metric. Metrics are reviewed to ensure that they reflect the intent correctly. For example, Nortel was removed from the TSX after dropping over 98% in under a year since it skewed the true perspective of market value.
While it’s never nice to make fun of a number – teasing values from a set of data is required to create and discern a metric. Simply stating that there are 12 art dealers in an area is irrelevant without context. Comparing this number to the amount of art dealers in other areas creates a metric. Metrics can be found in almost everything that can be measured on a scale, but occasionally, data can be misused or malformed. For example: Age, by itself, is not a metric. Average age, however, can be. Determining what is and what’s not a metric can be difficult. Data that can be categorised and not comparable, such as eye colour, cannot become a useful metric. But, the number of people with green eyes, however, can be, since it is a value that can be compared to other sets of data.
Independent and Dependent Metrics
An independent variable is a value that sits by itself and describes a topic without requiring background knowledge. An example of this can be employment or store vacancies. While comparable to values within other areas, this kind of metric does not need outside information to create meaning. 8% vacancy is just that, 8 out of 100 stores are empty.
Dependent variables, however, require an immediate comparison or a ratio. A metric that shows the percentile of ethnic stores holds value by itself, but a metric that compares this percentile to the number of ethnic people within an area can depict a much different message. While Corso Italia has a large Italian-centric marketplace, the surrounding area has very little Italian population in comparison to other ethnicities. This, as a metric, can show that it is a destination area, rather than a locally serviced area.
Baselines help place metrics into perspective. Often, this is the average value of all data, minus the area under review. In the example of store vacancies, the city average at around 8.8% would be a baseline. What is important to note about baselines is that while they depict an average, the average may not be ideal. Unemployment, overall at 10%, is a baseline, but at 9%, while better, is still too high. This is where thresholds come in.
Metrics can have thresholds. Average income, for instance has lower threshold, called the LICO, where being below this amount indicates that someone may have trouble making ends meet in their specific area. Age, on the other hand, may have an upper threshold, where above a certain amount may mean that there are less people who are willing to be employed or require differing needs. Or you’re dead or close to it.
Some metrics have dual thresholds: an upper and a lower. The Bank of Canada insists on keeping inflation between 1% and 3%; above would be too high for consumers to make debt payments and employers to match or meet wages demands. Below, investment in businesses will falter and growth – the driving force of capitalism – will whither. A practical example of this is store vacancies. Above a certain point can cause breakage in the streetscape, a negative reflection that can spiral out of control. Below a certain point, however will raise average rents as demand outstrips supply, placing pressure on existing businesses to compete for space with increasing rates.
Metrics are fluid, much like a bubble in a construction level; forces on either side tilt the metric forward and backward. Determining the scope of time required to capture a trend is sometimes difficult, specifically when limited by the availability of data, but understanding and depicting this can change the direction of policy.
A median income above the city average (or baseline) can be a positive for an area, but if that metric is trending downward, specifically over a set period of time, might mean a shifting demographic that will influence a business. Alternatively, an increasing median income might force businesses that currently exist under pressure to either change the quality of service or be forced out by new market forces that wish to gain benefits from the traction. (Starbucks, *ahem*)
Not only should a metric show trending, but it should also show the velocity of said trend. Leslieville, for example, in the past 12 years has seen the average price of housing increase from 155,000 to 450,000 – an increase by a factor of three. In an area once dotted with drop-in clinics, derelict stores and pawnshops, this pressure has changed the retail landscape dramatically. Alternatively, Forest Hill (along Eglinton Way) has seen an increase in housing prices, but at a lesser extent, not changing the retail market that dramatically.
After determining the thresholds and velocity of trends, a pattern emerges: the sweet-spot. This is where conditions are ideal. To use the inflation example again, 2% is a renowned sweet-spot: the economy is growing – expanding wealth – but not at such a level that it can become oppressive to industry and where wages can be adjusted by business to match easily. This allows sober reflection on direction and time to change accordingly. The ideal condition for the economy is not just growth, but measurable and stable growth.
Sweet-spots, much like everything else in a metric, are fluid – in as such, they move with trends. Levelling outside factors, aiming for a sweet-spot or maintaining one that has been achieved is the ultimate condition for a business to obtain.
While predominate classification of a metric is often as simple as: below-scope, at-scope (or sweet-spot) and above-scope; quite often ranges can be defined in a more normalised manner. As in measuring household income with quintiles, the definition of ranges can help provide context to a metric.
For example, defining a metric on a specific ethnicity may be concluded as such:
Each one of these ranges can help define an area, specifically when combined with the velocity of change and have a wide range of implications to either planning policy or businesses in the area, specifically when opposing metrics are competing
Metrics are not stand-alone; they can be combined to garner meaning. The sum of specific metrics aid in defining an area’s unique cache and can be leveraged to build more success or used to discover opportunities and threats.
On occasion, the combination of metrics can define an area, providing it with a persona and indicating specific qualities that it can control or aim for. These metrics can show where they are strong, where they are growing and where there are deficiencies.
Not all metrics can be combined, but often sets of metrics can be compared to create a profile that can be compared either directly or indirectly to other sets.
Developing a Profile
Much like an online dating site, metrics are used to develop a profile. You might think of this as a supra-metric where the addition of metrics combine to make a new metric that fits within a specific threshold. An area that requires input from outside customers can be defined via different metrics: ethnicity, as in Corso Italia or Gerrard with little or low similar demographics; clustering of retail, such as college street’s computer row; or low residential parameters to maintain the level of stores, such as downtown areas (although, with condo development, these numbers are trending in the opposite.)
Why This Is Important
Combining and comparing metrics, either individually or in sets, can help us define the qualities that make a BIA successful, or alternatively, unsuccessful. This will help categorise the type of BIA that best suits (ethnocentric, localised, clustered business, etc…) but also show the trends that may be affecting the area.
From this, we can describe planning remedies or policies that can help either deter the onset of change or propel an area to a desired sweet-spot. The marriage of policy to metrics is important to ensure that efforts are directed in the most effective and efficient manner possible, removing overlap or conflicting ideas and perhaps developing new ideas that can be applied elsewhere.
Do all areas possess the same metrics?
No. But, the data garnered from all areas help build the total list of metrics that apply to all areas. If this is an area that possesses little ethnicity or clustered stores, a metric may not be necessary – but, its weight in the context of the city does affect the outcome of dependent variables elsewhere.
Can an area be a part of two differing sets of metrics?
Definitely. An area that can be classified as a ethnic cluster could also possess a retail cluster. Kensington Market can be classified as a multicultural cluster and a food services cluster. These are defined by sets of metrics that are non-opposing, but when viewed in the scope of a supra-metric, can fall into two categories – or more.