Common Submittal Management Issues

Submittal problems are often diagnosed as diligence problems, but they rarely are.

The instinct is to conclude that somebody was not paying attention. The project engineer should have chased it. The subcontractor should have submitted earlier. The architect should have reviewed faster. Every time that same diagnosis happens, the same solution is given, which is more meetings and more reminders. The result is always a slight improvement that lasts for at most two months.

The actual causes are structural. Review capacity is finite and gets consumed in the wrong order. Submission deadlines in specifications are measured against a clock that has nothing to do with the building. Nobody is incentivized to submit early, and several parties are incentivized to submit everything at once. Approvals transfer risk in ways most people misunderstand. And the entire process contains a category of document that nothing in the workflow ever chases, because it requires no approval.

This article works through twelve failure modes. For each: the mechanism that produces it, the symptom you actually observe, what it costs, and the intervention that addresses the cause rather than the symptom. It closes with a diagnostic table, because most teams experiencing submittal trouble are treating the wrong issue.

Why the Diligence Diagnosis Fails

Consider a project where submittals are consistently late to the design team. The obvious reading is that subcontractors are not submitting on time.

Now look at the mechanism. The specification says submittals are due within thirty days of subcontract execution. The subcontractor submits on day twenty-eight, which is early against his obligation. The contractor forwards it on day thirty. The design team returns it in fifteen working days, requiring revision. The revised submittal comes back on day sixty, is returned approved on day eighty-one, and fabrication releases. The equipment has a thirty-week lead time and was needed on site in twenty-two weeks.

Nobody was late. Every party met or beat their contractual obligation. The building still waited eight weeks, because the only date anyone was measuring against was the wrong one.

That is the shape of most submittal failures. The process ran correctly and produced a bad result, which means the process was designed wrong rather than executed poorly. Chasing people harder does not fix a design problem.

Table 1. Symptom, Popular Diagnosis, and Actual Cause

Objectives

• Recognize each failure mode by its mechanism rather than by its symptom
• Intervene at the cause, since symptom-level interventions decay
• Understand which failures are the contractor’s to fix and which require the owner
• Diagnose correctly, because most teams are treating the wrong issue
• Build the small number of measurements that make each failure visible before it costs anything

Issue 1: The Incomplete Register

Mechanism. The register is built from a prior project’s template, or from Part 1 submittals articles alone. Requirements specified in quality assurance articles, source and field quality control articles, Division 01 sections, delegated design triggers, and addenda are never entered.

Symptom you observe. Nothing, for a long time. That is what makes this the worst one. Then a document is requested that nobody registered, usually near closeout or during an inspection.

Cost. Variable and occasionally severe. A missing mockup requirement discovered after fabrication should have started is a schedule event. A missing deferred submittal discovered by an inspector can stop installation.

Why it persists. Extracting every requirement means reading four article types across every section of a project manual that may run 1,200 pages, plus Division 01. Nobody completes that inside a buyout window, so templates get used, and the differences are invisible by definition.

Intervention. Audit rather than assume. Take three technical sections plus all of Division 01, extract every requirement from every article type, and compare against the register. The miss rate tells you what the log is worth. Most template-derived registers come back sixty to seventy-five percent complete, with the shortfall concentrated in mockups, qualification statements, commissioning documentation, and closeout items.

Issue 2: Specification-Clock Scheduling

Mechanism. Submit-by dates are taken from the specification, which states deadlines relative to subcontract execution or notice to proceed. Required-on-site dates come from the construction schedule. The two clocks are unrelated, and only one of them describes the building.

Symptom. Items that are on time by every internal measure and late by the only measure that matters. Procurement pressure with no identifiable late party.

Cost. The largest single category of submittal-driven schedule loss in commercial construction, and it lands on long-lead equipment where recovery options are expensive or absent.

Intervention. Backward scheduling. For every item, work from the required-on-site date through installation, delivery, shipping, fabrication, approval, review cycles, contractor screening, and subcontractor preparation. The result is the latest date the submittal can be submitted and, critically, the latest date the subcontract can be awarded.

Table 2. Issue 2, Where the Two Clocks Diverge

The last row matters, because it explains why the problem stays hidden. On most items the specification clock is conservative, and everything works. The failure is concentrated in the twenty or thirty items where it is not, and those are exactly the items that govern the schedule.

Where the intervention has to happen. During preconstruction, before buyout sequencing is fixed. On most commercial projects, three to six packages need to be awarded earlier than trade logic suggests, purely because of submittal and fabrication arithmetic. That is a buyout decision, and it cannot be made by whoever maintains the log.

Issue 3: The Front-Loaded Submission Flood

Mechanism. Specifications require submission within a fixed period after subcontract execution. Subcontractors comply, and because most subcontracts execute within a narrow window, several hundred submittals arrive at the design team over a few weeks. Review capacity is fixed. Queue depth explodes.

Symptom. Design team review durations balloon in months three through six, then normalize. Everyone concludes the consultant is under-resourced, and the consultant concludes the contractor is dumping.

Cost. Cycle times of six to ten weeks against a fifteen-working-day contractual duration, applied indiscriminately, which means long-lead items with no float wait behind stock items with ninety days of float.

Why did the diligence diagnosis fail here? Every party behaved correctly. The subcontractors met their deadlines. The contractor forwarded promptly. The consultant reviewed at a normal rate. The system was designed to produce a flood.

Intervention. Two things, and the second matters more.

Sequence submissions deliberately rather than accepting them as they arrive. A submittal schedule that spreads submissions by need date rather than by contractual deadline flattens the queue without anyone working harder.

Then prioritize the queue explicitly. Tell the design team which items have float under fifteen working days and ask for those first. Most consultants will accommodate a prioritized queue and cannot accommodate an undifferentiated one, because they have no basis for triage. Withholding priority information and then complaining about cycle time is a common and self-defeating pattern.

Issue 4: The Review Bottleneck

Mechanism. Distinct from Issue 3, though frequently confused with it. Here the consultant genuinely exceeds contractual durations independent of loading, because of resourcing, because reviews queue behind design work on other projects, or because a single individual is a single point of failure.

Symptom. Cycle times consistently over the contractual duration across the whole project, not concentrated in a period. Often concentrated in one discipline.

Cost. Compounds on every item, and it is the failure most likely to become a delay claim.

Intervention. Data and contract, in that order.

Report average cycle time by discipline monthly, against the contractual duration. Not total counts, not open items, cycle time by discipline. Consultants respond to being measured in a way they do not respond to being asked.

Then give written notice on the day each duration expires. Not at week four, not in a meeting. The day it expires. Contemporaneous documentation is what supports a delay position later, and it costs nothing at the time.

Then escalate to the owner, because the design consultants are under contract to the owner rather than to you. A persistent bottleneck is an owner problem with an owner remedy, and presenting it early with three months of cycle time data is far more effective than presenting it late as part of a claim.

Table 3. Issue 4, Distinguishing Loading From Capacity

Diagnosing this pair correctly matters, because the remedies are unrelated and applying the wrong one wastes months.

Issue 5: The Resubmittal Spiral

Mechanism. An item goes through three submissions and three rejections. Each cycle costs two review durations plus handling on both sides.

Symptom. Items reaching revision three or four while everyone treats each cycle as an isolated event.

Cost. On a fifteen-working-day review duration, three cycles is nine weeks of elapsed time consumed by one item.

The diagnostic insight. Repeated rejection is rarely a persistence problem in the submitter. It is one of four things, and each has a different remedy:

1. The subcontractor does not understand the requirement. Remedy: a conversation, usually fifteen minutes.
2. The specification is genuinely ambiguous. Remedy: an RFI, and the resubmittal is blocked until it closes.
3. The reviewer is applying a standard the documents do not state. Remedy: ask the reviewer to cite the requirement. This is more common than people expect, and raising it politely usually resolves it.
4. It is an unacknowledged substitution attempt. Remedy: return it and require the substitution procedure.

Intervention. Treat the third revision as a meeting trigger rather than another cycle. Bring together the subcontractor, contractor, and reviewer and determine which of the four causes is the issue. Most spirals close on the next cycle once the actual disagreement is named.

Also, separate contractor returns from design team revisions in your numbering. A submittal returned at intake should not consume a revision number, because conflating the two makes the first-pass approval rate unmeasurable and that metric is the main lever on this whole issue.

Issue 6: Unflagged Deviation and Disguised Substitution

Mechanism. Most general conditions require the submitter to identify deviations from the specification in writing. Most submitters do not, some through oversight and some deliberately. A deviation that reaches the design team unflagged may be approved by inattention, at which point the departure has arguably been accepted.

The related failure is a substitution submitted as product data. Division 01 sets a substitution procedure with a deadline, required comparative data, and a decision authority. A different product submitted as ordinary product data bypasses all of it.

Symptom. Approved work rejected at inspection or questioned during commissioning. Or a substitution discovered after the product was ordered.

Cost. High and awkward. Unwinding an approved deviation is a negotiation in which your position is weak, because your own process approved it.

Intervention. Compare the submitted product against the specification at intake, and flag deviations yourself when the submitter has not. This takes five minutes on most submittals. Where a different product has been submitted, return it and require the substitution procedure rather than forwarding it.

The second part of the intervention is a routing rule: any submittal carrying an identified deviation goes to the design team with a contractor cover note stating the deviation explicitly. That single practice converts a silent risk into a documented decision.

Issue 7: The Ball-in-Court Black Hole

Mechanism. An item has no recorded owner, or the owner field is stale. Nobody is chasing it because nobody knows it is theirs. Days accumulate invisibly.

Symptom. Items surfacing in a status meeting having aged forty or sixty days with no explanation, and no one able to say where they sat.

Cost. Pure waste, and it is entirely avoidable with two fields.

Intervention. Two fields and one report. Ball in court, days aged in current position, and a weekly report sorted by float to need. Not sorted by date received, which is how most logs report and which puts a trivial item with ninety days of float above a critical item with four.

The prioritization point deserves emphasis because it is counterintuitive to a filing mentality. Chronological order is fair and irrelevant. The working list should be sorted by consequence.

Issue 8: Circular Dependency With RFIs

Mechanism. A submittal is held pending an RFI answer. The RFI cannot be answered until the design team sees a submittal. Both logs show an item in progress. Neither shows that the two are waiting on each other.

Symptom. Two items that never move, each apparently active, discovered when somebody traces one of them manually.

Cost. Weeks, occasionally months, and the loop is invisible unless both logs cross-reference.

Intervention. A cross-reference field in both logs, and a rule: escalate the dependency, not the dependent item. Chasing a submittal that is legitimately held on an unanswered RFI accomplishes nothing and irritates the subcontractor. Chasing the RFI moves both.

Then add a weekly check for items held on a dependency more than ten days, which is where loops become visible.

Issue 9: Conditional Approval Ambiguity

Mechanism. A submittal returns with a disposition permitting work to proceed subject to a condition, or with comments requiring a response. The work proceeds. The condition is never closed.

Symptom. Discovered at inspection or commissioning, when the condition turns out to have mattered.

Cost. Variable, and the position is poor, because the contractor proceeded on a conditional approval and treated it as unconditional.

Intervention. Treat conditional approvals as requiring closure before release. Practically, that means a log field distinguishing conditionally approved from approved, and a rule that release for fabrication does not occur while the condition is open. Where the condition is minor and the schedule cannot wait, get the closure in writing rather than assuming it.

Issue 10: Approval Mistaken for Validation

Mechanism. An approved shop drawing contains a wrong dimension, a wrong quantity, or a coordination conflict. The contractor proceeds on the basis that the design team approved it.

Most general conditions state that the design team reviews for conformance with the design concept and the contract documents, and that the contractor remains responsible for field dimensions, quantities, fabrication means and methods, and coordination with other trades. An approval does not transfer those responsibilities.

Symptom. A dispute in which the contractor cites an approval and the design team cites the review limitation clause, and the clause usually wins.

Cost. The full cost of the error, plus the cost of having built more of it than you would have if you had caught it yourself.

Intervention. Contractor review at intake with the correct scope: completeness, coordination with other trades’ approved submittals, dimensional consistency with field conditions where field verification was required, and deviation identification. That is a screening review, not a design review, and it is exactly what the contractor is positioned and obligated to perform.

The related discipline concerns approvals returned as “approved as noted.” Those notes are instructions, and they occasionally conflict with the specification, with another trade’s approved submittal, or with a field condition. Reading them rather than filing them is the difference between an approval and a problem.

Table 4. Issue 10, Where Responsibility Actually Sits

Issue 11: Approval Without Distribution

Mechanism. A submittal is approved and the approval is recorded. The field never receives the current version, or receives it and continues using an earlier one because the earlier one is still in the trailer.

Symptom. Work installed per a superseded submittal. The correct version existed the whole time.

Cost. Rework, and it is entirely a distribution failure rather than an approval failure, which is why it is so frustrating.

Intervention. Three things. A single controlled location for current approved submittals, with superseded revisions marked void rather than removed silently. Distribution with acknowledgment recorded, so currency is verified rather than assumed. And a review of current submittal status with each trade before it mobilizes, which catches the crews who arrived with a package from three months ago.

Also notify procurement on approval. An approval that does not trigger a purchase order or a delivery confirmation has changed nothing about when material arrives.

Issue 12: The Closeout Crunch

Mechanism. Closeout submittals are specified in Part 1 closeout articles and Division 01 closeout, O&M, training, and commissioning sections. They are mostly informational, so nothing in the workflow stops when they are absent. And they are due at substantial completion, when the crews who could produce them have demobilized.

Symptom. The final six weeks consumed by chasing warranties, O&M data, record drawings, attic stock, and training records from twenty subcontractors, under retention pressure, with facilities staff waiting.

Cost. Retention disputes, delayed final payment, poor-quality documentation the owner will live with for thirty years, and relationship damage with subcontractors you want on the next project.

Intervention. Three changes, none of which requires anyone to work harder.

Register closeout submittals during setup, extracted from the correct articles rather than remembered at the end.

Derive their due dates backward from substantial completion rather than discovering them at it, and start collection at roughly seventy-five percent completion.

Tie specific closeout items to specific progress payments rather than to final payment alone. Retention is the wrong lever because it arrives after the people who could help have left the project. A payment application at eighty percent completion is a lever that works.

Table 5. Issue 12, Closeout Items and When to Start

Diagnostic: Which Issue Do You Actually Have?

Most teams experiencing submittal trouble are intervening on the wrong issue. This table maps observable symptoms to likely causes.

Table 6. Symptom to Diagnosis

The last row is worth dwelling on. A contractor return rate near zero looks like efficiency and means submissions are being forwarded unscreened. The rejections still happen, at the design team, at the cost of a full review cycle each. A healthy return rate is somewhere between ten and twenty-five percent.

Stakeholders and Who Owns Each Remedy

Table 7. Issue Ownership

Ten of the twelve are fixable inside the contractor’s organization without anyone’s permission. Only Issue 4 genuinely requires the owner, and only three require design team cooperation. That distribution is worth knowing, because submittal problems tend to be discussed as though they are somebody else’s fault, and mostly they are not.

Required Documentation

Table 8. Documents Required to Address These Issues

Technology Integration

Table 9. Which Issues Technology Addresses

Two observations. Platforms are strong on Issues 7, 11, and 4’s measurement, which are workflow and tracking problems. They are indifferent to Issues 1 and 2, which are the two that cause the most damage. And Issues 2, 3, and 10 are not technology problems at all: they are a calculation, a sequencing decision, and a contractual misunderstanding.

AI-Assisted Opportunities

Two of the twelve are document volume problems, and they happen to include the worst one.

Issue 1 requires reading four article types across every section of the project manual plus Division 01, which nobody completes in a buyout window. Issue 6 requires comparing each submission against the governing specification requirements item by item, including identifying deviations the submitter did not flag, which on a large submittal is careful reading that gets done fast or not at all.

Platforms that index construction documents address both. iFieldSmart AI, for example, extracts submittal requirements across every section and article with citations to the source, and checks submitted documents against the governing specification requirements to flag missing or non-compliant content. That converts register construction from sampling to coverage and gives intake screening a baseline rather than a memory.

Table 10. AI Use Case Matrix by Issue

Implementation

1. Run the register audit this week. Three sections plus Division 01. It takes a few hours, and it tells you whether anything else you do with the log means anything.
2. Run the backward calculation on long-lead items, and let the result change buyout sequencing. Issues 1 and 2 together account for most of the damage in this article.
3. Start measuring four things: first-pass approval rate, cycle time by discipline, count of items with float under fifteen days, and contractor return rate. Those four diagnose eight of the twelve issues.
4. Institute intake screening with written deficiency returns, and keep returns out of the revision counter.
5. Sequence submissions by need date and give the design team a prioritized queue. This addresses Issue 3 without requiring anyone else to change.
6. Add ball-in-court and days-aged fields and report weekly sorted by float to need.
7. Cross-reference the submittal and RFI logs and escalate dependencies rather than dependent items.
8. Read the review limitation clause in your general conditions and make sure your project engineers understand what an approval does and does not do.
9. Register closeout submittals at setup, start collection at seventy-five percent, and tie items to progress payments.

Table 11. Implementation Roadmap

Best Practices

Table 12. Practices That Prevent These Issues

Common Mistakes in Addressing These Issues

Table 13. Interventions That Do Not Work

How These Issues Appear on Real Projects

Class A office tower, 340,000 square feet. Issues 2 and 3 together. Curtain wall needed fabrication release eleven weeks earlier than the buyout schedule contemplated, and the submission flood in months four through six had review durations running at seven weeks against a fifteen-working-day requirement. The remedy was two-part, and neither part involved chasing anyone: move the glazing award forward, and hand the design team a queue prioritized by float.

Regional hospital expansion, 190 beds, occupied campus. Issue 6. A submitted air handling unit differed from the specified unit in filtration configuration, and the deviation was not identified by the submitter. On healthcare work, filtration is a licensing matter. Caught in a five-minute comparison at intake; approved by inattention, it would have surfaced during infection control commissioning with occupancy exposed.

Hyperscale data center, 60 MW. Issue 2 in its most severe form. Switchgear and transformer lead times put required award dates before the design was complete enough to buy from, which is not a submittal problem at all but a procurement strategy problem that only becomes visible when somebody runs the backward calculation during preconstruction. The response was an early release package with a defined change mechanism.

Advanced manufacturing facility, food grade. Issue 8. Process equipment submittals were held pending RFI answers about utility connection points, and the RFIs could not be answered until the vendor submittals arrived. Two logs, both showing active items, neither moving. Visible only once the cross-reference was added, and resolved by escalating the vendor document release rather than either dependent item.

University Science Building, 11 floors. Issue 5. Three laboratory casework items reached a fourth revision, and the intervention meeting revealed cause three: the reviewer was applying a finish standard the specification did not state. Two of the three closed on the next cycle once the requirement was clarified by RFI. Nine weeks had been spent on cycles before anyone asked what the actual disagreement was.

Multifamily podium, 240 units. Issue 12. Twenty-two subcontractors demobilizing progressively as buildings turned over, with warranties, O&M data, and attic stock outstanding. Deriving closeout dates backward and tying specific items to progress payments converted a retention fight into routine collection.

Highway interchange, design-bid-build. Issue 1, in the form that costs money rather than time. Certified material test reports were conditions of acceptance and payment, with requirements sitting in measurement and payment sections that the register never mined. Missing certifications meant unpaid work rather than delayed work, which is a different and equally unwelcome failure.

Historic warehouse conversion to hotel, 140 keys. Issues 1 and 2 compounding. Three sections required mockups approved before fabrication release of custom millwork and window assemblies. The mockup requirements were in Part 1 quality assurance articles and were not registered, and the backward calculation had not accounted for the additional approval gate. Discovering both at once cost seven weeks on the finish sequence.

Frequently Asked Questions

Which of these issues costs the most?

Issue 2, specification-clock scheduling, in aggregate, because it lands on long-lead equipment where recovery options are expensive or nonexistent. Issue 1 is the most dangerous per occurrence because it is unrecoverable. Issues 3 and 4 together cause the most visible friction and are the most commonly misdiagnosed as each other.

How do we tell whether our review problem is loading or capacity?

Plot cycle time against queue depth over the project. If cycle time rises and falls with volume, it is loading, and you control the remedy through sequencing. If cycle time is flat and consistently over the contractual duration regardless of volume, it is capacity, and the remedy is contractual notice followed by owner escalation. Table 3 sets out the other distinguishing signals.

Is it really the contractor’s job to fix a design team review bottleneck?

Not to fix it, but to document it correctly and escalate it to the party who can. The design consultants are under contract to the owner. What the contractor owns is the loading side, which is often half the problem, and the documentation, which determines whether a delay position holds later. Written notice on the day each duration expires costs nothing and is what makes the difference.

Why is a low contractor return rate a bad sign?

Because it means submissions are being forwarded without screening. The rejections still occur at the design team, and each one costs a full review cycle on both sides rather than an hour of screening. Ten to twenty-five percent is healthy. Near zero means the screening step is not happening, which drives Issues 5, 6, and 10.

What is the fastest improvement available?

Sorting the working list by float to need rather than by date received. It costs nothing, requires no cooperation from anyone, and immediately puts effort where consequence is. Second fastest is written notice on the day a review duration expires, which also costs nothing and materially improves your contractual position.

How do we stop deviations from slipping through?

Compare the submitted product against the specification at intake. Five minutes on most submittals. Where a deviation exists and was not identified, flag it yourself and forward with a cover note stating it explicitly. That converts a silent risk into a documented decision, and it protects you regardless of how the design team responds.

Does an approved shop drawing protect us if it turns out to be wrong?

Partly, and less than most people assume. Most general conditions state the design team reviews for conformance with the design concept and contract documents, while the contractor remains responsible for field dimensions, quantities, means and methods, and trade coordination. Read your specific clause. Approval offers real protection on design conformance and essentially none on the categories in Table 4.

How early should closeout collection start?

Around seventy-five percent completion for most items, and record drawings should be maintained continuously from mobilization because they cannot be reconstructed accurately afterward. The mechanism that makes it work is tying specific closeout items to specific progress payments, because retention leverage arrives after the crews who could produce the documents have left.

We already use a construction management platform. Why do we still have these problems?

Because platforms address workflow and are indifferent to register completeness and to backward scheduling, which are Issues 1 and 2 and the two most damaging. A platform gives you excellent control over whatever list you entered. If that list is missing forty percent of its requirements and has no need dates, the platform will manage it smoothly toward a bad outcome, and the smoothness makes the problem harder to see.

How do we handle a subcontractor who consistently submits late?

First verify they are actually late against a need-based date rather than early against a specification date, because the two frequently disagree. If they are genuinely late, escalate at the subcontractor’s management level rather than item by item, and bring the pattern rather than one instance. Where several subcontractors are all late, the dates are probably the problem rather than the subcontractors.

Which metrics actually indicate whether the process is healthy?

Four. First-pass approval rate, cycle time by discipline against the contractual duration, count of items with float to need under fifteen working days, and contractor return rate. Between them, they diagnose eight of the twelve issues in this article. Open item counts and total submittal counts are operationally useful and say almost nothing about health.

Can these issues be fixed mid-project?

Ten of the twelve, yes, and several within a week. The register audit, the four metrics, the float-based sorting, intake screening, ball-in-court fields, and the RFI cross-reference are all mid-project changes requiring nobody’s permission. Issue 2 is the one that cannot be fixed late, because buyout sequencing decisions already made cannot be unmade, though the calculation still tells you which items to expedite.

Expert Recommendations

Audit the register before doing anything else. If it is materially incomplete, every other improvement is being applied to the wrong list, and you will not find out for months.

Run the backward calculation on long-lead items even if buyout is already complete. It will not undo award decisions, and it will tell you which three items to expedite now rather than which twelve to explain later.

Start reporting four metrics monthly, including the contractor return rate. Between them they diagnose most of this article, and a low return rate is the single clearest signal that screening is not happening.

Hand the design team a prioritized queue with float information. Most consultants will accommodate priority and cannot triage an undifferentiated flood. Withholding the information and then complaining about cycle time is common and self-defeating.

Brief your project engineers on the review limitation clause in your general conditions. Most of them believe an approval offers more protection than it does, and that belief is what makes Issue 10 expensive.

Treat the third revision as a meeting. Nine weeks spent on cycles before anyone asks what the actual disagreement is happens on most projects, and it is entirely avoidable.

Tie closeout submittals to progress payments this project, not the next one. It is a single conversation with your project manager and it removes the most predictable crisis in construction administration.

Where This Leaves You

Twelve failure modes, and ten of them are fixable inside your own organization without anyone’s cooperation. That is an unusually favorable position, and it does not match how submittal problems are usually discussed, which is as something subcontractors and design teams are doing to you.

The two that matter most are also the two least likely to be recognized as problems, because neither produces a symptom until it produces a consequence. An incomplete register reports nothing. Specification-clock scheduling reports everything as on time. Both are diagnosable in a day and both are far cheaper to find in month one than in month fifteen.

Run the register audit and the backward calculation on the project you are on now. One will tell you what your log is missing. The other will probably tell you something uncomfortable about an item you have not thought about yet, which is exactly the point.