Why a single number misleads
Search for deck costs and you will find national averages assembled from data across the entire country. Those figures blend a ground-level platform in Alabama with an elevated walkout deck in eastern Idaho, and the result describes neither. No one publishes a Pocatello bid book either.
What follows is a band you can plan with, then the anatomy of a real quote. Pocatello pushes cost in ways a national average misses: 36-inch footings against 24 inches farther west, snow loading that drives heavier framing, and rocky soil that makes the dig harder. A written local bid that states footing depth is the number you build from.
The four variables
1. Height above grade — the biggest single factor
This matters more than square footage. A ground-level platform under 30 inches needs no guardrails, short posts, minimal bracing, and often no permit. The same deck at walkout-basement height — six to nine feet, extremely common in Pocatello — needs guardrails throughout, a full stair run, tall posts with lateral bracing, and a structurally consequential ledger.
Two decks of identical square footage at different heights are not comparable projects. If your budget is tight and your grade allows it, going lower saves more than any material choice will.
2. Footings and the dig
36 inches, on every post, in Bannock County's rocky soil. This is real labour and real concrete, it scales with the number of posts, and it is invisible once poured — which makes it the easiest thing for a low bid to shave.
Ask every builder to state footing depth and diameter in writing. A bid noticeably below the others is very often a bid that plans to dig shallower.
3. Material
Pressure-treated pine is the budget standard. Cedar sits above it. Composite is substantially more up front and eliminates refinishing. Railing is its own decision and a large share of the total — composite railing in particular costs considerably more than wood, which is why a composite deck surface with pressure-treated railing and a composite top rail is a common compromise. More on material choice here.
4. Complexity
Multi-level construction, angled or curved sections, built-in seating or planters, integrated lighting, and roofs or pergolas all add cost disproportionately to their size. A covered deck in particular becomes a structural design question because of snow load and 115 mph design wind — it is not a small upgrade to a deck project, it is a second project.
Planning ranges
These bands are for a simple ground-level rectangle: short posts, little or no railing, no roof, and no demolition of an old deck. They are installed prices — labor and materials — drawn from Idaho and Treasure Valley compilations published for 2026. A Boise remodeler's published platform prices sit inside the same bands. Use them to tell a real bid from a fantasy. They are a planning range for this valley, and the bid in your hand supersedes them.
That is about $20–$40 a square foot in pressure-treated wood and $35–$60 in capped composite, for the platform. Cedar sits between the two. Cellular PVC, the TimberTech AZEK class of board, usually prices above the composite band. Ask for it as its own line.
A 12 by 16 is 192 square feet, so read it against the 200-square-foot band. A 16 by 20 is 320, so it sits with the 300. The bands are close together on purpose. Adding a hundred square feet moves the price less than lifting the same deck off the ground.
That card is the 200-square-foot platform, plus about 44 feet of wood guard, plus one stair run and the taller posts. It is not a second published average. Six to nine feet up is the ordinary deck on the Bench, and on plenty of houses elsewhere in town. The footings still have to reach 36 inches, in rocky ground, and there are more of them doing more work. A composite guard is still extra, on the per-foot line below. If the grade allows a lower deck, dropping the height saves more than switching from composite back to wood.
Left out of every figure above: the permit, often $50–$300; any railing past a basic wood guard; and extra digging when a footing hits rock. Rock does not change the 36-inch depth. It changes how long the hole takes.
Railing, by the foot
Railing is not inside the ground-level bands. The walkout card already includes a basic wood guard. Anything past that is its own line. Regional installed prices, for planning:
- Wood guard, about $30–$50 per linear foot.
- Composite or aluminum, about $45–$90 per linear foot.
- Cable, about $60–$120 per linear foot. The posts cost more than the cable.
A raised 12 by 16 with the house side open has about 44 feet of guard before the stairs. Near the middle of these per-foot bands that is roughly $1,800 in wood and $3,000 in composite or aluminum. The wood figure is already inside the walkout card above. The stair rail on a walkout adds more feet. What the guard has to do.
Line items to ask for separately
Ask each builder to break the bid out. Not so you can price the pieces yourself, but because a bid missing one has told you something.
| Line item | Why it matters | Impact |
|---|---|---|
| Footings and excavation | Depth, diameter and count. The invisible cost and the usual place a cheap bid cuts. | Major |
| Framing package | Beam sizing, joist spacing, hangers and connectors. Spacing must suit the decking product. | Major |
| Ledger and flashing | Small line, largest safety consequence. Should be specified, not assumed. | Moderate |
| Decking material | Product line and grade, not just "composite". | Major |
| Railing | Frequently a surprising share of total. Priced per linear foot. | Major |
| Stairs | Scales with height. A walkout deck's stair run is a real cost. | Moderate |
| Permit and inspections | Commonly $50–$300 across Idaho. Pocatello uses a valuation schedule. See the permits page. | Minor |
| Demolition and disposal | Only if replacing an existing deck. | Minor |
When one quote comes in well under the others, it is almost never because that builder found efficiencies. In this market it is usually one of three things: shallower footings, lighter framing than snow load warrants, or no permit. All three are invisible on completion and expensive in year six. Ask the low bidder to specify footing depth, joist spacing and who pulls the permit — the answers usually explain the gap.
Where the money is well spent
- Footing depth. Non-negotiable. Everything above it depends on it.
- Ledger detailing and flashing. Small money, catastrophic failure mode.
- Framing one step heavier than minimum. Cheap at build time given local snow loads.
- Permit and inspections. The only independent check on work you will never see again.
Where you can reasonably economise
- Build lower if grade allows. Biggest saving available.
- Pressure-treated instead of composite, if you will genuinely maintain it.
- Wood railing with a composite top rail rather than full composite railing.
- Simple rectangle. Angles and curves cost real money for aesthetics.
- Build in the off season. Late summer and autumn availability is better and bids are often keener than in the May–June rush.
Common questions
How many quotes should I get?
Three. One tells you nothing, two gives you no way to spot the outlier, and beyond three you are mostly burning time. Give each builder the same brief so the bids are comparable.
Is a deck a good investment?
Deck additions recoup a reasonable share of cost at resale nationally, but treat that as a secondary benefit. The primary return is that you use it. Building a deck you do not want because a national report says it recoups well is a poor reason.
Should I pay a deposit?
A deposit is normal — materials have to be bought. A large up-front payment before any work starts is not. Structure payments against milestones: deposit, footings complete and inspected, framing complete and inspected, final on completion. Never pay in full before the final inspection passes.
Does a bigger deck cost proportionally more?
Not linearly. Fixed costs — mobilisation, the ledger, the stair run, the permit — spread across more square footage, so cost per square foot generally falls as size rises. Height and complexity move the number far more than area does.