Updated September 2026 · Written and maintained by the Progression Agency strategy team
Backlog means work that has been committed but not yet completed. That single definition covers six quite different business situations, and two of them point in opposite directions: a rising sales backlog is usually good news, while a rising support backlog is always bad news. This page defines each sense, explains why the count is the least useful number you can report about one, sets out how to work a backlog down in an order that actually holds, and covers the words backlog gets confused with — queue, pipeline, roadmap and technical debt.
The short answerA backlog is accumulated work that is committed but not yet done. In sales and manufacturing it is orders won and not yet delivered, and it is an asset. In operations, support and product it is work owed and not yet reached, and it is a liability. The number to report is not the count — it is the aging: how old the oldest item is, and what the median age is. Two teams reporting an identical backlog count can be in completely different situations, and only the age distribution tells them apart. And a backlog whose inflow exceeds its throughput cannot be cleared by working harder, which is the single most common mistake made with one.
Backlog meaning, in one sentence
Answer first: a backlog is the accumulated body of work that has been committed to but not yet completed. The word carries no judgment on its own — whether a backlog is good or bad depends entirely on which of the six senses below you mean.
Where the word comes from
The term originally referred to a large log placed at the back of a fire to burn slowly, and it carried the sense of a reserve held back. The modern business usage keeps the reserve idea and loses the fire: a stock of work held, waiting, not yet consumed.
Backlogs meaning in the plural
Answer first: identical. Backlogs, backlog, backlogs means, backlog means — these are the same word being searched with different grammar, not different concepts. The plural is used when several separate backlogs exist, which in most organizations they do: a sales backlog, a support backlog and a product backlog can all exist at once and behave completely differently.
The six senses of backlog
Answer first: sales or order backlog, operations backlog, support backlog, product backlog, financial or billing backlog, and technical debt. They share the definition and almost nothing else, including whether a rise is good news.
| Sense | What is waiting | Rising is | Who watches it | The health measure |
|---|---|---|---|---|
| Sales / order backlog | Orders won, not yet delivered | Usually good | Finance and leadership | Deliverability against capacity |
| Operations backlog | Work owed, not yet performed | Bad | Operations | Aging and inflow versus throughput |
| Support backlog | Tickets received, not yet resolved | Always bad | Support leadership | Age of oldest open ticket |
| Product backlog | Features considered, not built | Neutral to bad | Product teams | Whether anything ever leaves it |
| Billing backlog | Work delivered, not yet invoiced | Bad | Finance | Days between delivery and invoice |
| Technical debt | Quality work deferred | Bad, slowly | Engineering | Rate of accumulation versus repayment |
The conclusion: when somebody says ‘the backlog is up’, the first question is which one. A rising order backlog and a rising support backlog are opposite pieces of news and the same three words describe both.
Sales and order backlog
Answer first: orders a company has won but not yet delivered or recognized as revenue. In manufacturing, construction and enterprise software this is a genuine forward indicator and it is frequently reported to investors, because it says something about revenue already secured.
When a sales backlog stops being good news
When it grows because delivery cannot keep up rather than because sales are strong. A large order backlog with lengthening delivery times is a capacity problem wearing a growth costume, and customers experience it as lateness regardless of how it appears on a chart.
Operations backlog
Answer first: work the organization owes and has not reached. Fulfillment, installations, maintenance, processing. Unlike sales backlog, this is unambiguously a liability, and its cost compounds because delayed work generates its own follow-up work — chasing, apologizing, re-scheduling.
Support backlog
Answer first: tickets received and not resolved. This is the sense where aging matters most, because customer tolerance decays sharply with time and an old unresolved ticket generates more contact rather than less.
Product backlog
Answer first: a list of features, fixes and ideas a product team might build. It is the sense most often misused, because in many organizations it has become a place items go to avoid somebody having to say no to them. That is not a backlog; it is an archive with optimistic labeling.
Billing backlog
Answer first: work delivered but not yet invoiced. It is quietly expensive, because every day between delivery and invoice is a day of financing somebody else’s business, and it is usually invisible until somebody looks for it deliberately.
Technical debt
Answer first: quality and maintenance work deferred in favor of speed. It is a specific kind of backlog rather than a synonym for one, and its distinguishing feature is that it accrues interest — deferred work makes future work slower, which generates more deferral.
Why aging matters more than count
Answer first: because two teams with an identical backlog count can be in completely different situations, and only the age distribution distinguishes them. Report the oldest item and the median age; the count on its own is close to meaningless.
| Measure | What it tells you | What it hides | Report it? |
|---|---|---|---|
| Total count | Volume waiting | Everything about urgency | Only alongside aging |
| Age of oldest item | Whether anything is stuck | Volume | Always |
| Median age | Whether the tail is typical or exceptional | The extremes | Always |
| Items over 90 days | The size of the stuck tail | Recent flow | Always |
| Inflow per period | Whether the problem is arriving faster | Capacity | Always |
| Throughput per period | What you can actually clear | Demand | Always |
| Inflow minus throughput | Whether it is growing or shrinking | Composition | Always — this is the key number |
The conclusion: the last row is the number that decides whether a backlog is solvable with effort. If arrivals exceed completions, no amount of working harder clears it — the arithmetic simply moves the date at which it becomes unmanageable.
How to work a backlog down
Answer first: measure aging, stop the inflow, separate items by cause, delete the dead ones, sequence what remains by one consistent rule, then fix the intake. In that order — most teams start at sequencing and wonder why it does not hold.
Step one: measure aging, not count
Before anything else, produce the age distribution. It converts ‘we have four hundred items’ into ‘we have three hundred and forty items under a week old and sixty that have been open for more than a year’, which are two entirely different problems requiring two entirely different responses.
Step two: stop the inflow
If inflow exceeds throughput, the backlog grows regardless of effort. Reducing inflow means either preventing the work arriving — fixing the cause of the tickets, tightening what gets accepted — or increasing throughput permanently. A temporary push clears the visible pile and the pile returns.
Step three: separate by cause
Four categories, and only one of them needs capacity: blocked on something external, waiting on a named person, genuinely not started, and should be deleted. Teams routinely apply capacity to all four, which is why backlog pushes exhaust people and achieve less than expected.
Step four: delete the dead
Answer first: deletion is a legitimate completion and treating it as failure is the main reason backlogs grow forever. An item nobody will ever do costs attention every single time somebody reads the list, and that cost is paid weekly for years.
Step five: sequence by one rule
Oldest-first or highest-value-first. Both are defensible; alternating between them is not, because it produces a list where the ordering logic is unknowable and everybody negotiates their own item upward.
Step six: fix the intake
Otherwise the same backlog reassembles within a quarter. The intake fix is usually either a definition of what gets accepted, a required piece of information at submission, or removing the underlying cause that generates the work.
Backlog, queue, pipeline, roadmap and WIP
Answer first: these are five different things and they get used interchangeably. A queue is ordered, a pipeline is not yet committed, a roadmap has dates, work in progress is being done now, and a backlog is committed work waiting.
| Term | Committed? | Ordered? | Being worked? | Has dates? |
|---|---|---|---|---|
| Backlog | Yes | Not necessarily | No | No |
| Queue | Yes | Yes, by rule | No | No |
| Pipeline | No — still being won | Loosely, by stage | No | Sometimes forecast |
| Roadmap | Partially | By theme | No | Yes |
| Work in progress | Yes | N/A | Yes | Usually |
| Wish list | No | No | No | No |
The conclusion: the distinction that matters most in practice is backlog versus wish list. A backlog contains work somebody intends to do. A wish list contains work somebody did not want to refuse. Mixing the two makes the backlog unusable for planning, which is exactly what has happened in most product organizations.
Backlog by industry
Answer first: the word means something specific and different in each of these, and the metric that matters differs with it.
| Industry | Backlog means | Reported to | The number that matters |
|---|---|---|---|
| Manufacturing | Orders received, not yet shipped | Investors, as a forward indicator | Backlog months of production |
| Construction | Contracted work not yet built | Investors and lenders | Backlog in months of revenue |
| Enterprise software | Contracted revenue not yet recognized | Finance and investors | Remaining performance obligations |
| Support and service | Tickets not yet resolved | Operations leadership | Age of oldest open item |
| Product development | Features and fixes not yet built | Internal only | Whether anything ever leaves it |
| Legal and professional | Matters not yet concluded | Partners | Capacity utilization |
| Recruitment | Roles open, not yet filled | Leadership | Time to fill |
| Healthcare | Patients waiting for a procedure | Regulators and boards | Longest wait, not average wait |
Note the last row in particular. In contexts where waiting has a human cost, the longest wait is the honest figure and the average wait is the flattering one — which is why the average is the one usually quoted.
Is a backlog good or bad?
Answer first: it depends on whether the backlog represents demand you have won or work you owe. Won-and-undelivered is an asset with a capacity risk attached; owed-and-unreached is a liability with a customer-experience cost attached.
- A sales backlog is good until delivery times lengthen, at which point it becomes a capacity problem that customers experience as lateness.
- An operations backlog is bad and gets worse than linearly, because delayed work generates chasing, apologizing and rescheduling work of its own.
- A support backlog is unambiguously bad, and its cost is measured in customers lost rather than in tickets outstanding.
- A product backlog is neutral if items leave it and bad if they do not. A product backlog where nothing is ever deleted is a record of indecision.
- Technical debt is bad slowly, which is why it is deferred, and which is why it eventually stops being deferrable.
The life of a backlog item
The stage nobody measures is ‘blocked’. Most backlog items spend the majority of their life waiting on something or someone rather than waiting for capacity, and a team that measures only capacity will keep adding people to a problem that is not about people.
A complete question set on backlogs
Answer first: the questions people actually ask about backlogs, each answered directly. If you are trying to make sense of one in your own organization, these are the answers to have to hand.
Definition and scope
Is a backlog the same as being behind schedule?
No. A backlog can be entirely on schedule — an order backlog with delivery dates being met is simply committed work waiting its turn. Being behind means the committed dates are not being met, which is a different problem.
Can a backlog be zero?
In support and operations, briefly, and it usually indicates either very low demand or excess capacity. In sales a zero order backlog means nothing is sold ahead, which for most businesses is worse news than a large one.
Is a backlog an asset or a liability?
An order backlog is closer to an asset — revenue secured, not yet delivered. An operations or support backlog is a liability — work owed, not yet done. The word alone does not tell you which.
Does a backlog appear on financial statements?
Not as a line item in the way inventory does, but related concepts do — remaining performance obligations under revenue recognition standards, for instance. Companies frequently discuss backlog in commentary alongside statements rather than within them.
Measurement
What is backlog burn rate?
The rate at which backlog is converted into completed work, usually expressed per period. Compared against inflow, it tells you whether the backlog is growing or shrinking, which is the only question that matters at the start.
What is backlog coverage?
How long the current backlog would sustain output at current throughput — frequently expressed in months. In manufacturing and construction it is one of the standard forward indicators.
Should I measure backlog in items or in effort?
Effort, wherever items vary widely in size. Four hundred items where one takes a day and another takes a month is not a meaningful count, and effort-weighted backlog is the honest version.
How do I spot a backlog that is being gamed?
Look for items closed and immediately reopened, items reclassified into a different queue, and a count that falls while the age of the oldest item rises. All three reduce the reported number without reducing the work.
Product backlogs specifically
How big should a product backlog be?
Small enough that somebody can read all of it. Once it exceeds that, items at the bottom are effectively invisible and the list has stopped functioning as a decision tool.
What is backlog grooming or refinement?
The regular practice of reviewing items, clarifying them, re-ordering them and deleting the ones that will never be built. The deletion half is the part most often skipped and the part that keeps the list usable.
Should old product backlog items be deleted or archived?
Deleted, in practice. Archiving usually means moving them somewhere nobody looks, which is deletion with extra steps and a false sense of preservation. If an idea matters it will be raised again.
Why do product backlogs grow forever?
Because adding an item is a way of not saying no. It resolves the immediate conversation at no apparent cost, and the cost is paid later by everybody who has to read the list.
Practical management
Who should own the backlog?
One person for the list as a whole, plus a named owner for each item. Shared ownership of a backlog produces a list nobody prunes, because pruning somebody else’s item feels presumptuous.
Should I cap the size of a backlog?
It is a useful forcing device. A hard cap means adding something requires removing something, which converts the vague intention to prioritize into an actual decision made at the moment of addition.
What is a reasonable aging target?
It depends entirely on the sense and the sector, so set your own from your own data rather than borrowing one. A useful starting discipline is that nothing should be older than the period in which you review the backlog.
How do I explain a growing backlog to leadership?
With the inflow and throughput figures rather than the count. ‘Arrivals exceeded completions by sixteen a week for eleven weeks’ is a diagnosis; ‘the backlog is up’ is a complaint, and it invites the wrong response.
Common mistakes with backlogs
| Mistake | Why it costs | Instead |
|---|---|---|
| Reporting the count only | Hides whether anything is stuck | Report oldest and median age too |
| Adding effort without stopping inflow | Clears the pile; the pile returns | Fix inflow first |
| Never deleting anything | Dead items cost attention weekly | Delete, do not archive indefinitely |
| Using the backlog as a wish list | Makes it useless for planning | Say no, or record it elsewhere |
| Alternating sequencing rules | Everybody negotiates their item upward | One rule, held |
| Averaging the wait | Hides the stuck tail | Report the longest wait |
| Treating blocked and undone the same | Applies capacity where it cannot help | Separate by cause first |
| No owner per item | Items age without anybody responsible | An owner and a decision date each |
What causes backlogs in the first place
Answer first: four causes, and only one of them is a capacity problem. Demand exceeding capacity, work arriving faster than it can be triaged, items blocked on something external, and work that should never have been accepted. Adding people solves the first and none of the others.
| Cause | What it looks like | The fix | What does not work |
|---|---|---|---|
| Genuine capacity shortfall | Steady inflow, throughput below it, few blocked items | More capacity, permanently | A temporary push |
| Triage failure | Large volume of unclassified items | A triage step with an owner | Working the list in arrival order |
| Blocking | Items old but untouched, not for lack of effort | Chase the blocker, not the item | Adding capacity |
| Accepting work that should be refused | Backlog full of items nobody will do | A definition of what gets accepted | Prioritizing harder |
| Rework | The same item returning repeatedly | Fix the quality problem upstream | Closing it faster |
| Seasonality | Predictable peaks | Plan capacity around the peak | Treating each peak as a surprise |
The conclusion: diagnose before you resource. Most backlog interventions add capacity because that is the visible lever, and in three of the six rows above capacity is not the constraint.
Backlogs and customer experience
Answer first: customers do not experience your backlog count; they experience their own wait. That is why the longest wait is the honest metric and the average is the flattering one, and why communication about a wait changes the experience of it more than a marginal reduction in the wait itself.
- Tell people where they are. An acknowledged wait is tolerated far better than a silent one, and it reduces the chasing contacts that inflate the backlog further.
- Give a date, not a platitude. ‘Within ten working days’ is usable; ‘as soon as possible’ generates a follow-up.
- Update when the date changes, before it passes rather than after.
- Handle the oldest items specifically. The stuck tail is where complaints and lost customers come from, and it is invisible in the average.
- Measure chase contacts. They are a direct readout of how well you are communicating, and they are pure additional inflow caused by your own silence.
Reporting a backlog honestly
Answer first: count, oldest item, median age, items over ninety days, inflow, throughput, and the difference between the last two. Seven numbers, one line each, and together they tell the whole story that the count alone conceals.
Support backlog, week ending [DATE]
Open items: 412
Oldest open item: 214 days
Median age: 4 days
Items over 90 days: 38
Inflow this week: 260
Throughput this week: 244
Net change: +16
Read that example and the situation is immediately legible: current flow is roughly healthy, there is a stuck tail of thirty-eight items that has been ignored, and the backlog is growing slowly because inflow exceeds throughput by sixteen a week. None of that is visible in ‘412 open items’, which is what most reports contain.
Related reading on the surrounding measurement discipline: how to calculate a success rate and what retention means.
Updated August 2026. The worked examples and distributions on this page are illustrative figures chosen to show the shape of the problem, not measured findings from any organization, and they are labeled as such where they appear.
Reporting a backlog nobody can act on?
Send us how it is currently measured. We will tell you what the count is hiding, whether the inflow or the throughput is the real problem, and how to report it so the answer is legible in one line.
Measurement and operations, from the people who publish the platforms
Publicly available talks from Google Ads, Think with Google, Ad Age, HubSpot, Ahrefs and Neil Patel on measurement and reporting discipline — the habits that make any operational number trustworthy. None of these are ours; each is credited to its channel by name and upload date, every identifier was verified live before publication, and each tile loads its player only when you click it.
Getting found in search
AI, AEO and what is changing
Paid media and lead generation
Websites and design
Choosing and working with an agency
- Marketing agencies in NYC
- Digital marketing agency NYC
- Digital marketing in Atlanta
- Digital marketing in Connecticut
- Digital marketing in Oklahoma City
- Nonprofit marketing agency
- Pharmaceutical marketing agency
- MSP marketing agency
- App development agency
- Growth tools for startups
- How to choose a digital marketing agency
- What does a marketing agency do?
- What does a digital marketing agency do?
- Marketing agency pricing
- What does retention mean?
- How to calculate success rate
- Best marketing agency
- Branding services for small businesses
- Fashion consulting
- Triadic color schemes
- The 8 principles of design
- How to write a brochure
- TikTok Wrapped: what actually exists
- Full-service marketing agency
- Marketing agencies in Philadelphia
- Agencies in Portland, Oregon
- Boutique marketing agency
- Marketing agency for small business
- Digital marketing services
- Digital marketing agency in New Jersey
- Fractional CMO services
- What is a fractional CMO?
- About Progression Agency
- Startup marketing agency
- Marketing agency in Tampa
- Marketing agency in San Diego
- Digital marketing in Florida
- Marketing agency in Nashville
- Marketing agency in Portland
- Client testimonials
- Contact us
- Marketing agency near me
- Online reputation case studies
- Digital marketing agency, Dallas
- Orlando digital marketing agency
- Advertising agencies in Westchester
- Water damage restoration leads
- General contracting leads
- Home inspection leads
- Spray foam insulation leads
- Mold remediation leads
- Digital marketing agency in Toronto
- Digital marketing internships
- Marketing consulting firms
- Marketing agency in Austin
- Advertising agency, Columbus
- Advertising agency, Charlotte
- Digital transformation consulting
- Ethos in advertising
- Sales promotion examples
- Experiential marketing agency
- Marketing strategy
- Small business marketing
- Brand awareness for professional services
- Healthcare strategy consulting
- Digital healthcare marketing
Social, content and brand
By industry and by situation
Backlogging: meaning, and what the verb form signals
Backlogging is the verb form, and it carries a meaning the noun does not. To backlog something is to place it deliberately into the queue rather than act on it now — a decision, not an accident. In software and operations teams the word is often used defensively: backlogging a request is how a team says no without saying no, and a backlog that grows faster than it is worked is a record of decisions nobody wanted to make out loud. In inventory and manufacturing the same word describes orders accepted but not yet fulfilled, where backlogging is a deliberate policy — you take the order now and produce later. The distinction worth holding onto is that a backlog is a state and backlogging is a choice, which is why the second is the more revealing word when you hear it in a status meeting.
Data center finance, and MRD
Two terms from adjacent industries that show up in backlog conversations.
Data center finance has become a distinct capital market. Data center debt is raised against the long contracted revenue of the leases, which is what makes the asset class attractive to lenders: a hyperscale tenant on a fifteen-year lease produces a bond-like cash flow. Data center securitization packages those lease revenues into securities sold to investors, which is why facilities are increasingly built by specialist developers with access to that capital rather than by the operators themselves.
MRD meaning in manufacturing is material requirements date — the date by which a component must be on hand for production to run to plan. MRD meaning in business more broadly is sometimes market requirements document, the artifact stating what a market needs before a product requirements document states what will be built. The two are unrelated and are distinguished by context: a supply chain conversation means the date, a product conversation means the document.
Backlog versus roadmap
A backlog is an ordered list of work that could be done; a roadmap is a statement of intent about outcomes. Confusing the two produces a roadmap nobody can commit to and a backlog nobody can prioritise.
A backlog that is never cut is not a backlog
Items that have sat unranked for a year are not waiting, they are declined. Removing them makes the list honest and shortens every prioritisation meeting.
Estimation is for sequencing, not for promising
Relative sizing helps decide order. Converting it into a date is where teams lose credibility, because the estimate never carried that information.
Frequently asked questions
What does backlog mean?
What does backlogs mean in the plural?
Is a backlog good or bad?
What is a sales or order backlog?
What is a product backlog?
What is the difference between a backlog and a queue?
What is the difference between a backlog and a pipeline?
What is the difference between a backlog and a roadmap?
Is technical debt a backlog?
Why is backlog aging more important than the count?
What backlog metrics should I report?
How do I reduce a backlog?
Why can’t I clear a backlog by working harder?
Should I delete items from a backlog?
How do I categorize backlog items?
What is a backlog in manufacturing?
What is a backlog in construction?
What is a support backlog and how should it be measured?
What is a billing backlog?
Why is the average wait misleading?
What is ‘blocked’ and why does it matter?
Should every backlog item have an owner?
How often should a backlog be reviewed?
What is the single biggest mistake with backlogs?
What does backlogging mean?
Get a free marketing proposal
Tell us what you are trying to grow and we will come back with a plan, not a pitch deck. Same-day reply on weekdays.