How to Audit and Measure Your Directory Listings

The six articles before this one have followed directory submission from its definition through to the act itself: what it is, whether it is worth doing, how it serves local search, how to choose directories, how to submit, and how to resist the instinct to submit too widely. This final article takes up what happens afterwards. A directory listing, once created, does not look after itself; it drifts out of accordance with reality as the business changes, and its value, whatever that value turns out to be, is invisible until it is measured. The activity therefore has a tail, and the tail has two parts that this article keeps carefully distinct: the audit, which keeps the listings accurate, and the measurement, which establishes what the listings actually produce. An owner who has read the series this far has learned to make good listings; this article is about not letting them quietly decay, and about finally answering, with evidence rather than argument, the question the whole series has circled — whether a given listing was worth making.

The standard note on sources applies. Peer-reviewed research is cited by author and year and listed at the end, and claims that rest on search-marketing practice rather than formal research are identified as such.

Why auditing and measuring are the other half of the activity

It is tempting to regard directory submission as finished once the submissions are made, and an owner who regards it so will have done, in fact, only the first half of the activity. The second half is required by two facts established earlier in the series. The first is that listings decay. The third article showed that a business’s details are not fixed — it moves, changes its telephone number, alters its hours, revises its name — and that a listing created once and never revisited drifts, with time, into stating things that are no longer true. A listing left unaudited is therefore not a stable asset; it is a slowly spoiling one, and the spoilage, as that article showed, both misleads the people who encounter the listing and undermines the corroboration that consistent listings provide. Auditing is the work of arresting that decay, and it is not optional, because the alternative to auditing is not a listing that stays good but a listing that gets worse.

The second fact is that the value of directory submission is, until it is measured, unknown. The second article delivered a verdict on what directory submission can do, but a verdict about the activity in general is not the same as knowledge about a particular business’s particular listings, and that article was explicit that the general guidance should be overridden, once it exists, by evidence about the specific case. Measurement is how that evidence is produced. Without it, an owner is left maintaining a set of listings on faith, unable to say which of them earns its place; with it, the owner can tell. Auditing and measuring are therefore not housekeeping appended to the real activity. They are the half of the activity that turns a set of submissions into a managed asset — accurate because it is audited, and understood because it is measured — and an owner who omits them has not done a lighter version of directory submission but an incomplete one.

Two distinct tasks: the audit and the measurement

Before describing either task it is worth separating them firmly, because they are routinely run together and the conflation weakens both. The audit and the measurement ask different questions, about different things, and an owner who keeps them apart will do each better. The audit asks: are the listings correct? It is concerned with the state of the listings themselves — whether each one states the business’s details accurately, in the canonical form, in the right category, without duplication. It is an inward-looking check, comparing the listings against a standard the owner controls, and its output is a list of corrections to make. The audit can be performed whether or not the listings have produced anything, because it is about their accuracy, not their results.

The measurement asks a different question: are the listings producing anything? It is concerned not with the state of the listings but with their effect — with whether people are reaching the business through them, and which listings are responsible. It is an outward-looking assessment, examining evidence about what the listings have caused, and its output is an understanding of which directories are earning their place. A listing can be in perfect order, accurate and consistent and well categorized, and still produce nothing measurable, just as a listing might produce referral visits while containing a stale detail; the two questions genuinely come apart. An owner needs both answers, because they support different decisions — the audit tells the owner what to fix, the measurement tells the owner what to keep — and the rest of this article treats them in turn, the audit first because a listing should be made correct before its results are judged.

The directory audit: keeping the listings accurate

The audit is the periodic, systematic check that every one of a business’s directory listings still states the truth, in the agreed form. It is not difficult work, but it is work that cannot be done at all without a preliminary the next part describes, and it is work that an owner will not do reliably unless it is given a fixed occasion rather than left to be remembered. The three parts below set out how the audit is built, what it checks, and how often it should happen.

Building the inventory of listings

An audit is a check of a known set of things, and so the first requirement is to know the set: an owner cannot audit listings whose existence they have not recorded. The inventory is simply a maintained record of every directory in which the business is listed — where the listing is, and, ideally, when it was last checked. For a business that has followed this series from the start, building the inventory is trivial, because the fifth article’s submission checklist ended with the instruction to record each listing as it was made; the inventory is just that record, kept up. For a business that has been listed over years by many hands, some of them not the owner’s, the inventory has to be assembled, and assembling it involves a deliberate search for listings the owner did not create — the listings, mentioned in the third article, that directories sometimes generate from other sources without the business’s involvement. Until the inventory exists the audit cannot be systematic, because the owner will be checking the listings they happen to remember rather than the listings that exist, and the listings most likely to be wrong are precisely the ones created by others and therefore the ones least likely to be remembered. The inventory is the unglamorous foundation of the whole audit, and it is the part most often skipped.

Checking each listing against the canonical form

With the inventory in hand, the audit itself is a comparison. For each listing in the inventory, the owner looks at the listing as it is currently published and checks it against the canonical form of the business’s details — the single agreed version of the name, address, and telephone number that the third article insisted be settled before any submission. The check is for any divergence: a detail that has fallen out of date because the business changed and the listing did not, a detail that was entered in a non-canonical form at submission, a description that has become inaccurate, a category that no longer fits because the business’s offering has shifted. Each divergence found is a correction to make, and the correction is made in the directory itself, bringing the listing back into accordance with the canonical form. The purpose, as the third article established at length, is consistency: the audit exists to ensure that the business’s many listings continue to agree with one another and with reality, because it is agreement that corroborates the business’s details and disagreement that undermines them. The check is mechanical and not hard; its difficulty is entirely a matter of remembering to do it, which is why the cadence below matters.

Finding duplicates and stray listings

One part of the audit deserves separate attention because it addresses a problem the simple per-listing check does not catch: the duplicate. A business can come to have two listings in the same directory — one created by the owner, another generated by the directory from another source, or two created at different times by different hands — and a duplicate is a particular kind of harm, because the two listings will almost never agree exactly, so a duplicate is also, automatically, an inconsistency. The audit should therefore include a deliberate search for duplicates within each directory, and where one is found the right action is to consolidate: to claim the listings, keep one, correct it to the canonical form, and have the directory remove or merge the other. The same search should look for stray listings more broadly — listings in directories the owner never chose to be in, perhaps created automatically — and for each the owner should decide whether the directory is one worth being in, by the fourth article’s test, claiming and correcting the listing if it is and seeking its removal if it is not. This part of the audit is the one that most rewards the inventory-building described above, because duplicates and strays are, by definition, the listings an owner did not deliberately create and so would not otherwise think to check.

How often the audit should happen

An audit that is performed when an owner happens to think of it will be performed rarely and irregularly, which is to say it will not reliably catch the drift it exists to catch. The audit needs a cadence. Two occasions should trigger it. The first is the calendar: a full audit of the inventory, performed on a fixed schedule — for most businesses once or twice a year is a reasonable interval, frequent enough to catch drift before it has spread far and rare enough not to be burdensome. The second occasion is event-driven and is the more important of the two: any real change in the business’s details — a move, a new telephone number, a change of name or hours — should trigger an immediate, targeted update of every listing in the inventory, not at the next scheduled audit but at the time of the change. The reasoning is that the scheduled audit catches slow, unnoticed drift, while the event-driven update prevents the largest and most damaging divergences, the ones where a known change is simply not propagated. An owner who audits on a schedule and updates on every change will keep a listing set consistent with very little ongoing effort; an owner who does neither will find, at some later point, a set of listings that disagree in ways that have been quietly costing the business for months.

Measuring what the listings produce: referral traffic

The audit keeps the listings accurate; it says nothing about whether they are doing any good. That is the measurement’s task, and the measurement begins with the one benefit that can be observed directly. The second article identified referral traffic as the most concrete of directory submission’s surviving benefits, and concrete is the right word: when a person clicks the link in a directory listing and arrives at the business’s website, that arrival is recorded, and the ordinary analytics of a website distinguish visitors by where they came from. An owner can therefore see, for each directory in which the business is listed, how many visitors that directory sent over a given period — which directories sent many, which sent a few, and which sent none at all. This is real evidence, attributable to specific directories, and it is the firmest single input the measurement has.

Reading it well requires two cautions. The first is patience: a listing’s referral traffic is rarely substantial in its first weeks, and a directory should be given a period of some months before its referral figures are treated as meaningful, because a judgement made too early will mistake the absence of accumulated data for the absence of effect. The second caution is more important and is taken up at length in the section on measurement’s limits: referral traffic is a floor on a directory’s effect, not the whole of it. A person may see a business in a directory, not click the directory’s link, and instead reach the business another way — by searching its name, or by typing its address directly — and the analytics will credit that later channel, not the directory that actually prompted the visit. Referral traffic therefore undercounts a directory’s true contribution, sometimes considerably. This does not make it useless; it makes it a measurement to be read as a reliable lower bound rather than a complete account. A directory showing substantial referral traffic has demonstrably earned its place; a directory showing none has not demonstrated its worthlessness, only failed to demonstrate this particular benefit.

Measuring the benefits that are harder to see

Referral traffic is measurable because it leaves a record; the other surviving benefits of directory submission are harder to measure precisely because they do not, and honesty requires saying so plainly rather than pretending to a precision that is not available. Three benefits fall into this harder category. The local-search citation effect, examined in the third article, cannot be measured per listing at all: it is a corroboration effect, produced by the agreement of many listings together, and there is no way to isolate the contribution of one citation to it, because the contribution is not individual. The credibility benefit, the reassurance a listing in a vetted directory gives a cautious stranger, operates inside the mind of that stranger and leaves no trace an owner can read. And a good deal of discovery is genuinely untraceable: a person who finds a business in a directory and remembers it, returning later by some other route, has been served by the directory in a way no analytics will attribute to it.

This does not mean these benefits cannot be assessed at all; it means they must be assessed by proxy and by inference rather than measured directly. Two proxies are worth using. The first is the directory’s own analytics, where the directory provides them: many genuine directories can report how often a listing was viewed and how often it was acted upon within the directory, and these figures, while not the same as referral traffic to the website, are real evidence that the listing is being seen by the directory’s audience. The second proxy is the oldest and most undervalued measurement available to any business: asking customers how they found it. A simple, consistently asked question — recorded over time — will, in aggregate, reveal patterns that no analytics captures, including the role of directories in discovery that left no digital trace. Neither proxy yields a precise figure, and an owner should not pretend otherwise. But the honest position is not that the harder benefits are unmeasurable and therefore to be ignored; it is that they are to be assessed with rougher instruments, and that an owner who uses the rough instruments deliberately will know considerably more than one who measures only what is easy and assumes the rest is nothing.

Table 1. The benefits of a directory listing, and how each can be assessed

Turning measurement into decisions

Measurement is not an end in itself; its purpose is to inform decisions about which listings to keep, which to correct, and which to let go, and the measurement is wasted unless it is actually used to make those decisions. The decision the measurement supports is, in outline, a pruning and a strengthening. A directory that the measurement shows to be producing — sending referral traffic, registering activity in its own analytics, named by customers as the way they found the business — has demonstrated its worth, and an owner should keep its listing, maintain it carefully in the audit, and, where the directory offers worthwhile ways to improve the listing, consider doing so. A directory that the measurement shows to be producing nothing, across a fair period and across every proxy available, is a candidate for removal: the listing can be let go, and the small ongoing effort of auditing it redirected to the listings that earn it.

The decision must, however, be made with one crucial qualification, and it is the qualification that the distinction between the benefits forces. A directory should not be pruned on referral traffic alone, because, as the previous sections established, some genuine benefits leave no referral trace. A foundational or local directory whose role is chiefly citation and corroboration may show little or no referral traffic and still be doing exactly the job it is there to do, and an owner who prunes it for want of clicks has destroyed a citation to satisfy a measurement that was never the right measurement for that listing. The decision rule must therefore be benefit-aware: a listing whose intended benefit is referral traffic is fairly judged by referral traffic, while a listing whose intended benefit is citation is judged by whether the directory remains a genuine, relevant one, not by clicks it was never expected to send. Used with that qualification, measurement closes the loop the series has been building toward. The second article gave a general verdict and said it should be overridden by evidence about the specific case; the measurement is that evidence, and turning it into decisions is how an owner finally manages their directory listings by what is true of their own business rather than by general rules, this series’ own included.

The limits of measurement

Because this article has urged measurement, it must also be honest about how far measurement can be trusted, since an owner who over-trusts an incomplete measurement will make confident decisions on poor evidence. Three limits should be held in mind. The first, already established, is that referral traffic undercounts: it is a lower bound on a directory’s effect, and a directory that sends no measurable referral traffic has not been shown to be worthless, only to be unproven on that one dimension. An owner who treats zero referral traffic as proof of zero value will prune directories that were quietly doing real work. The second limit is that attribution is imperfect even where it exists: the path by which a person reaches a business is often longer and more tangled than analytics can represent, involving several touches of which the directory was one, and a measurement that assigns a visit to a single source simplifies a reality that was not single. The third limit is that the benefits genuinely differ in measurability, as the table set out, and that the easily measured benefit — referral traffic — is not the most important one; for many businesses the citation effect matters more, and it is precisely the benefit that cannot be measured per listing at all.

The honest conclusion is not that measurement should be distrusted or skipped — an owner who measures knows far more than one who does not — but that measurement should be held with the same calibrated confidence the rest of this series has recommended for every claim about directory submission. Measurement is good evidence, not perfect evidence; it should inform decisions, not dictate them mechanically; and where it is silent, its silence should be read as the absence of a particular kind of proof rather than as proof of absence. An owner who measures carefully, reads the measurement for what it is, and prunes cautiously rather than aggressively will manage a directory presence well. An owner who measures and then treats the numbers as the whole truth will, with the best intentions, make some decisions that the fuller picture would have reversed.

Concluding remarks

Auditing and measuring are the half of directory submission that turns a set of submissions into a managed asset, and an owner who omits them has done the activity incompletely. The two tasks are distinct and should be kept so: the audit asks whether the listings are accurate and keeps them so, through a maintained inventory, a periodic check against the canonical form, a deliberate search for duplicates, and a fixed cadence of scheduled and event-driven updates; the measurement asks whether the listings produce anything, through the directly observable evidence of referral traffic and the rougher proxies — the directory’s own analytics, and customers asked how they found the business — that the harder-to-see benefits require. Measurement turned into decisions closes the loop, letting an owner prune what does not earn its place and keep what does, provided the pruning is benefit-aware enough not to discard a citation for want of clicks; and measurement held within its limits, as good evidence rather than perfect evidence, keeps those decisions sound.

This article completes the series, and the series has a single argument running through all seven of its parts. Directory submission was, for a decade, a link-acquisition tactic, and as that tactic it is finished, deliberately ended and not coming back. What survives is older and quieter: directory submission is the careful, selective, accurate placement of a business into catalogues that genuine audiences consult, for the sake of discovery, of the citations that support local search, of modest referral traffic, of a contribution to the machine-readable record, and of a measure of credibility — followed by the audit that keeps the listings true and the measurement that establishes their worth. Done that way, by the few who do it deliberately rather than the many who still do it in bulk, directory submission is a small, sound, and genuinely useful part of how a business is found. It is not what its reputation says it is, in either direction: neither the dead relic nor the ranking lever, but a real and bounded activity that repays being done with care and repays nothing at all when it is not.

Future developments

The auditing and measurement of directory listings are likely to become both easier and more important, and the two movements reinforce each other. Easier, because the instruments are improving: the tools for assembling an inventory of a business’s listings, checking them for consistency across many directories, and detecting duplicates have been getting more capable, and as they improve, the audit that this article has presented as a modest discipline should become more modest still, less a manual chore and more a supervised, largely automated check. More important, because the value of accuracy is rising: as the systems that mediate discovery come to depend more heavily on corroborated, machine-readable records, the consistency that the audit maintains will matter more, and the cost of the drift it prevents will rise accordingly. Measurement is likely to improve more slowly and more partially, because its central limits — that referral traffic undercounts, that attribution is imperfect, that the citation effect cannot be isolated — are not limits of current tools but of the thing being measured, and better instruments will not dissolve them. The probable future, then, is one in which keeping listings accurate becomes easier and matters more, while knowing exactly what they are worth remains, as it is now, a matter of good evidence honestly read rather than of precise proof. An owner who has learned to audit on a cadence and to measure within the limits of what measurement can do is equipped for that future — and has, with this article, the last of what the series set out to provide.

References

Akerlof, G. A. (1970). The market for “lemons”: Quality uncertainty and the market mechanism. The Quarterly Journal of Economics, 84(3), 488–500.

Bakos, J. Y. (1997). Reducing buyer search costs: Implications for electronic marketplaces. Management Science, 43(12), 1676–1692.

Hagiu, A., & Wright, J. (2015). Multi-sided platforms. International Journal of Industrial Organization, 43, 162–174.

Stigler, G. J. (1961). The economics of information. Journal of Political Economy, 69(3), 213–225.